FINAL REPORT FOR THE ADMINISTRATIVE CONFERENCE OF THE UNITED STATES

AGENCY APPELLATE SYSTEMS

Christopher J. Walker

The Ohio State University Moritz College of Law

&

Matthew Lee Wiener Administrative Conference of the United States

This report was prepared for the consideration of the Administrative Conference of the United States. It does not necessarily reflect the views of the Conference (including its Council, committees, or members). Recommended Citation Christopher J. Walker & Matthew Lee Wiener, Agency Appellate Systems (Dec. 14, 2020) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/final-report-agency-appellate-systems.

AGENCY APPELLATE SYSTEMS Christopher J. Walker* & Matthew Lee Wiener**

TABLE OF CONTENTS TABLE OF CONTENTS ................................................................................................ I

LIST OF ACRONYMS ............................................................................................... III

INTRODUCTION ........................................................................................................ 1

I. BACKGROUND ON AGENCY APPELLATE REVIEW ................................................ 5

A. Coverage: Defining Appellate Review .................................................... 5

B. Structures of Agency Appellate Review ................................................. 7

C. Models and Objectives of Agency Appellate Review ............................. 9

D. Conference Recommendations and Model Appellate Rules ................ 16

E. Judicial Review ..................................................................................... 18

II. STUDY METHODOLOGY ................................................................................ 20

III. FINDINGS..................................................................................................... 22

A. Law of Appellate Review ...................................................................... 23

B. Decisionmakers ..................................................................................... 28

C. Selecting Cases for Appellate Review .................................................. 29

D. Decision-Making Process and Decisions .............................................. 32

E. Administration, Management, and Bureaucratic Oversight .............. 40

F. Public Disclosure and Transparency ................................................... 46

G. Common Challenges in Agency Appellate Programs .......................... 51

IV. RECOMMENDATIONS .................................................................................... 53

A. Objectives of Appellate Review ............................................................ 53

* John W. Bricker Professor of Law, The Ohio State University Moritz College of Law; Chair,

American Bar Association Section of Administrative Law and Regulatory Practice. ** Acting Chairman, Vice Chairman, and Executive Director, Administrative Conference of

the United States. For helpful comments at various stages of this study, we thank Michael Asimow, Kent Barnett, Emily Bremer, Reeve Bull, Blake Emerson, Jeremy Graboyes, Jeff Lubbers, David Marcus, and Gavin Young. For vital research assistance, we thank Andrew Bair, Bret Baker, Ruben Garza, Morgan Huff, Sam Lioi, Megan Porter, and Brittney Welch. We also thank the numerous agency officials who agreed to be interviewed or otherwise shared their expertise and wisdom on the subject.

ii

B. Procedures for Appellate Review ......................................................... 54

C. Case Selection for Appellate Review .................................................... 55

D. Appellate Decision-Making Processes and Decisions.......................... 56

E. Administration, Management, and Bureaucratic Oversight .............. 57

F. Public Disclosure and Transparency ................................................... 58

CONCLUSION ......................................................................................................... 60

APPENDIX A TEMPLATE FOR AGENCY-SPECIFIC OVERVIEWS .................................... I

APPENDIX B APPELLATE PROVISIONS OF MODEL ADJUDICATION RULES (RULES 400–450) ............................................................................................................. VIII

INDEX OF ONLINE APPENDICES C–N: AGENCY-SPECIFIC OVERVIEWS .................. XVI

iii

LIST OF ACRONYMS

(PROPER NAMES)

Each acronym is written out the first time it appears in the Report. For convenient reference, we have prepared the following list.

ACUS Administrative Conference of the United States ALJ Administrative Law Judge APA Administrative Procedure Act BIA Board of Immigration Appeals BVA Board of Veterans’ Appeals DOJ Department of Justice DOL Department of Labor EAB Environmental Appeals Board (Environmental Protection

Agency) EEOC Equal Employment Opportunity Commission EOIR Executive Office of Immigration Review EPA Environmental Protection Agency HHS Department of Health and Human Services MSPB Merit Systems Protection Board NLRB National Labor Relations Board OMHA Office of Medicare Hearings and Appeals (Department of

Health and Human Services) PTAB Patent Trial and Appeal Board PTO Patent and Trademark Office USCIS U.S. Citizenship and Immigration Services SEC Securities and Exchange Commission SSA Social Security Administration

INTRODUCTION

This Report returns to an important but little studied subject that the Administrative Conference of the United States (ACUS) last addressed nearly forty years ago: internal agency review (sometimes called administrative review)1 of hearing-level adjudicators’ decisions—or, as we call it, “agency appellate review.” In 1983, ACUS addressed the legal structures of appellate review.2 The main question it asked then, as it had over a decade earlier,3 is when and how agencies heads, if not constrained by statute, should delegate their authority to review the decisions of administrative law judges (ALJs).

The answers ACUS gave to that question (summarized in Part I.D) remain generally sound despite intervening changes in agency adjudication. We see no reason to revisit them. They should continue to inform Congress’s and agencies’ design of appellate programs, although we have recommended several refinements. That is true whether the programs review decisions resulting from proceedings governed by the formal hearing provisions of the APA4 (the exclusive subject of ACUS’s earlier inquiries) or decisions of non-APA proceedings of sufficiently similar formality.

Our focus in this Report lies elsewhere. We take the legal structures of appellate programs as they are now constituted and ask, among other important questions: How should the programs structure their decision-making processes? What cases should they review? Under what standards of review? What procedural rules should they use? What form should those rules take? How should they promulgate the rules? What form should their decisions take? (When, for instance, should decisions be designated precedential?) What extra-decisional activities might they undertake to improve the hearing-level decisions they review and also their own decisions? What bureaucratic

1 See MODEL ADJUDICATION RULES WORKING GROUP, ADMIN. CONF. OF THE U.S., MODEL ADJUDICATION RULES 75–85 (rev. 2018) [hereinafter MODEL RULES], https://www.acus.gov/ research-projects/model-adjudication-rules-2018-revisions.

2 Admin. Conf. of the U.S., Recommendation 83-3, Agency Structures for Review of Decisions of Presiding Officers Under the Administrative Procedure Act, 48 Fed. Reg. 57,461 (Dec. 15, 1983). That Recommendation was supported by an excellent report by Professor Ronald Cass. See Ronald Cass, Agency Review of Administrative Law Judges’ Decisions (1983) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/project-report-recommendation-83-3; see also Russell L. Weaver, Appellate Structures in Executive Departments and Agencies, 48 ADMIN. L. REV. 251 (1996); JEFFREY B. LITWAK, A GUIDE TO FEDERAL AGENCY ADJUDICATION, 2012 A.B.A. SEC. OF ADMIN. LAW & REG. PRAC. 110–18. Professor Russell Weaver’s study, like Professor Cass’s, was limited to review of ALJ decisions. But unlike the latter, Weaver’s study addressed only certain case-study agencies.

3 Admin. Conf. of the U.S., Recommendation 68-6, Delegation of Final Decisional Authority Subject to Discretionary Review by the Agency, 38 Fed. Reg. 19,783 (July 23, 1973). The associated report is James O. Freedman, Review Boards in the Administrative Process, 117 U. PA. L. REV. 546 (1969).

4 See 5 U.S.C. §§ 554, 556–557.

2

mechanisms might they employ to carry out their missions as efficiently and fairly as scarce resources allow? What information should they share with the public? It is important to emphasize up front that these questions, and all others we consider here, can usually be addressed by agencies as a matter of administrative choice. Few agencies are constrained by statute.

Appellate review government-wide is a vast subject to say the least, and our choice of questions to address has been necessarily selective. Our choice has been driven by consideration of what questions might be given general answers applicable to all or most appellate programs, but not so general as to be useless. Pragmatic considerations have played an important role. They include whether agencies are likely to find our suggestions worth serious consideration.

Some of our questions will need to be answered differently by different appellate programs. Answers may depend on the nature of the decisions a particular appellate program reviews, its caseload, the (often limited) resources available to it, and the objective(s) it serves.

We give particular attention to the last of these considerations: the objective of, or reason for, appellate review in any given program—or, more accurately in most cases, the objectives of review, since most programs have multiple, and sometimes competing, objectives. Of course, what is a best practice for a program whose main objective is simple error correction may not be the best practice for a program whose main objective is, say, formulating policy or achieving systemic consistency across a high volume of cases. The important thing, though, is that every agency appellate program should consider, decide, and disclose publicly what objective(s) it is serving. Agencies seldom do so. Our first recommendation is directed to remedying this shortcoming.

Some of the best practices we offer are cast at some level of generality—and necessarily so. There are hundreds of agency appellate programs. Each is different, often in significant ways. Procedural uniformity across programs—sometimes called trans-substantive procedure5—would be neither possible nor desirable in administrative adjudication,6 whatever its merits as an aspiration for federal-court rulewriters.7 Still less possible or desirable would be cross-system uniformity with respect to internal management and other processes.

We expect, though, that agencies will find practical use for even our most general recommendations. Some identify cross-cutting legal norms to which nearly every program should adhere, no matter what it adjudicates and how it does so. Others identify important foundational questions to which all agencies

5 See, e.g., David Marcus, Trans-Substantivity and the Process of American Law, 2013 B.Y.U. L. REV. 1191 (2014).

6 See, e.g., Matthew Lee Wiener, General Rules for Agency Adjudication?, REG. REV. (Oct. 29, 2018), https://www.theregreview.org/2018/10/29/wiener-general-rules-agency-adjudications.

7 See, e.g., Stephen B. Burbank, Pleading and the Dilemmas of Modern American Procedure, 93 JUDICATURE 109 (2009).

3

should attend in designing new and reviewing existing appellate programs, even if they must be answered differently for each such program. Of course, this study is no substitute for studies of specific agencies. ACUS has already undertaken a number of such studies;8 it may want to undertake additional ones.

There are a number of important subjects related to agency appellate review (other than those issues noted above) that lie outside the Report’s scope. Several bear mention. One is the quality of appellate decisions and of the decision-making process that affects appellate review. Measuring decisional quality would require specification of a standard for assessment—no simple matter9—and then a statistical analysis of a program’s decisions. Neither is possible in a broad study like this. Less possible still is an assessment of whether a program’s decision-making process effectively advances the agency’s objective(s) for appellate review.

A related subject we mention, but do not address, is bureaucratic quality-assurance review of hearing-level (or appellate) decision making. Quality-assurance programs may “operate,” as an excellent recent study notes, “as a more systematic replacement for a laborious . . . system of appellate review.”10 That may be especially true in high-volume adjudication systems in which accuracy and systemic consistency are especially important. In leaving quality-assurance programs largely unaddressed, we do not mean to foreclose the possibility that they may sometimes serve as an alternative or (more realistically) a supplement to conventional agency appellate review. We note in the Conclusion that ACUS may wish to consider a study on this important subject.11

There are, finally, two legal issues not addressed here yet meriting mention. The first concerns agency-appointed adjudicators who issue final agency decisions—that is, without further agency-head review—under delegations of authority from politically appointed agency heads or by statutory authorization. A serious constitutional question has recently arisen as to whether these

8 See Part I.D infra. 9 Even the most obvious standard—success on judicial review—is problematic. See Jonah

Gelbach & David Marcus, A Study of Social Security Litigation in the Federal Courts 71–77 (2016) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/report-study-social-security-litigation-federal-courts.

10 David Ames, Cassandra Handan-Nader, Daniel E. Ho & David Marcus, Due Process and Mass Adjudication: Crisis and Reform, 72 STAN. L. REV. 1, 76 (2020); see also Robert A. Kagan, Varieties of Bureaucratic Justice, in ADMINISTRATIVE LAW FROM THE INSIDE OUT: ESSAYS AND THEMES IN THE WORD OF JERRY MASHAW (Nicholas R. Parrillo ed., 2017) (noting “administrative audit system” as one alternative to “case-by-case appeal”).

11 ACUS addressed this subject in Recommendation 73-3, Quality Assurance Systems in the Adjudication of Claims of Entitlement to Benefits or Compensation, 38 Fed. Reg. 16,840 (June 8, 1973). The underlying report is Jerry L. Mashaw, Quality Assurance Systems in the Adjudication of Claims of Entitlement to Benefits or Compensation (1973) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/project-report-recommendation-73-3.

4

adjudicators are principal officers under the Constitution12 and hence must be appointed by the President (with Senate confirmation) rather than by other agency officials.13 A case pending before the Supreme Court raises the issue in one particular context.14 Our recommendations do not depend on how that case or any others are resolved, although some recommendations will be useful to agencies that decide to provide for agency-head review,15 whether for policy reasons or to remediate potential constitutional problems.

A second and related legal issue we do not address concerns the terms and conditions under which agencies employ appellate adjudicators who are not agency heads. The most contentious of them is what, if any, legal protections against removal agencies should accord them, above and beyond what the civil service laws may already provide.16 ACUS studied this issue several years ago, but could not reach consensus on how it should be resolved on a cross-agency basis for either hearing-level or appellate adjudicators.17 Although we do not revisit it here, we have recommended that agencies disclose certain information

12 U.S. CONST. art. II, § 2, cl. 2. 13 See Fleming v. Dep’t of Agric., No. 17-1246 (D.C. Cir.); see also Arthrex, Inc. v. Smith &

Nephew, Inc., 941 F.3d 1320 (Fed. Cir. 2019) (holding that agency-appointed administrative patent judges at the Patent and Trademark Office—who, by statute, issue final decisions—are unconstitutionally appointed principal officers), cert. granted, No. 19-1434 (Oct. 13, 2020). See generally Richard J. Pierce, Jr., The Collision Between the Constitution of the 1930s and the Constitution of 2020, YALE J. ON REG.: NOTICE & COMMENT (Dec. 18, 2019), https://www.yalejreg.com/nc/the-collision-between-the-constitution-of-the-1930s-and-the-constitution-of-2020-by-richard-j-pierce-jr; William Funk, Fallout from Lucia—The Unconstitutionality of the Environmental Appeals Board and Interior Board of Land Appeals, YALE J. ON REG.: NOTICE & COMMENT (Jan. 19, 2019), https://www.yalejreg.com/nc/fallout-from-lucia-the-unconstitutionality-of-the-environmental-appeals-board-and-interior-board-of-land-appeals-by-william-funk.

14 Arthrex, 941 F.3d at 1320, cert. granted, No. 19-1434 (Oct. 13, 2020); see also Christopher J. Walker, The Supreme Court Decides to Hear Arthrex Case to Consider Appointments Clause Challenge to Administrative Patent Judges, YALE J. ON REG.: NOTICE & COMMENT (Jan. 19, 2019), https://www.yalejreg.com/nc/the-supreme-court-decides-to-hear-arthrex-case-to-consider-appointments-clause-challenge-to-administrative-patent-judges/.

15 See U.S. Dep’t of Labor, Secretary’s Order No. 01-2020, Delegation of Authority and Assignment of Responsibility to the Administrative Review Board, 85 Fed. Reg. 13,186 (Mar. 6, 2020) (amending procedures to provide for the Secretary’s review of decisions of internal appellate bodies previously unreviewable).

16 See Kent Barnett et al., Non-ALJ Adjudicators in Federal Agencies: Status, Selection, Oversight, and Removal (2018) (report to the Admin. Conf. of the U.S.) [hereinafter Barnett, Non-ALJ Adjudicators], https://www.acus.gov/report/non-alj-adjudicators-federal-agencies-status-selection-oversight-and-removal. This report was published in part as Kent Barnett & Russell Wheeler, Non-ALJ Adjudicators in Federal Agencies: Status, Selection, Oversight, and Removal, 53 GA. L. REV. 1 (2019). Professor Barnett has since made a strong case for agencies to adopt regulations that would enhance the impartiality of adjudicators. See Kent Barnett, Regulating Impartiality in Agency Adjudication, 69 DUKE L.J. 1695 (2020).

17 See Admin. Conf. of the U.S., Administrative Judges, https://www.acus.gov/research-projects/administrative-judges (last visited Nov. 24, 2020).

5

about adjudicators’ terms and conditions of employment.18 That may overlap with recommendations arising from another pending ACUS project devoted exclusively to the disclosure of information about adjudicators.19

The Report proceeds as follows. Part I provides necessary background on agency appellate programs (their structures, objectives, and legal bases), prior ACUS recommendations on the subject, and judicial review of agency adjudication decisions. Part II explains our study methodology. Part III sets forth our main findings. Part IV details our recommendations for ACUS’s consideration. The Conclusion offers some possibilities for further ACUS study.

I. BACKGROUND ON AGENCY APPELLATE REVIEW

A. Coverage: Defining Agency Appellate Review Defining agency “appellate review” can be problematic, especially outside the

context of review of ALJ decisions. The term does not have a single accepted meaning in agency adjudication, as appellate review does in the federal courts. Numerous adjudication programs hear “appeals”—and may even have “appeal” in their names—but they would not normally be thought of as undertaking an appellate function. While these programs provide for review of a front-line official’s adjudication decision (rendered after an informal decision-making process),20 they usually do so through a de novo evidentiary proceeding that results in a “new” decision, which itself is then subject to higher-level agency review. (The evidentiary record before the front-line official may still play an important role in the proceedings.) The proceedings are trial-like, not appellate. Examples include the adjudication proceedings handled by HHS’s Office of Medicare Hearings and Appeals (OMHA)21 and the MSPB’s administrative judges in the exercise of the MSPB’s so-called “appellate jurisdiction.”22

At the same time, the term “appellate review” cannot be reserved exclusively for agency programs (of which there are many) that review adjudication decisions resulting from a full evidentiary hearing on the basis of a closed and

18 See Part III.F.6 infra. 19 Admin. Conf. of the U.S., Public Availability of Information About Agency Adjudicators,

https://www.acus.gov/research-projects/public-availability-information-about-agency-adjudicators (last visited Nov. 24, 2020).

20 As distinct from, say, a prosecutorial-type decision to issue a complaint. See MICHAEL ASIMOW, ADMIN. CONF. OF THE U.S., FEDERAL ADMINISTRATIVE ADJUDICATION OUTSIDE THE ADMINISTRATIVE PROCEDURE ACT 8–9 (2019) [hereinafter ASIMOW, ADJUDICATION SOURCEBOOK], https://www.acus.gov/publication/federal-administrative-adjudication-outside-administrative-procedure-act.

21 OMHA characterizes its decisions as “new.” U.S. Dep’t of Health & Hum. Servs., Level Three Appeals (OMHA), https://www.hhs.gov/about/agencies/omha/the-appeals-process/level-3/index.html (last visited Nov. 24, 2020).

22 See App. C (HHS); App. J (MSPB).

6

exclusive record of documentary evidence, transcripts, and dockets of the sort familiar in federal courts.23 One reason why is that any number of adjudicative offices that otherwise perform only traditional appellate functions sometimes hear new evidence on appeal in a way unknown to federal-court appellate practice. The SSA Appeals Council is but one example.24

Another reason is that some adjudicative offices that perform mostly trial-like functions also perform appellate functions. The trial and appellate functions can even be commingled. A notable example (though an outlier, to be sure) is offered by the Patent Trial and Appeals Board (PTAB) at the U.S. Patent and Trademark Office (PTO).25 It has both hearing-level and appellate jurisdiction. Even its original-jurisdiction adjudications have an appellate character. Its judges sit in three-member panels and issue final agency decisions, reviewing decisions by patent examiners. The decisions are subject to reconsideration, sometimes with additional judges added to the panel, by order of the PTO’s director. The director through a precedential opinion panel may designate panel decisions as precedential and hence binding on the program’s several-hundred judges in all future cases.26 Any number of other adjudication programs likewise defy easy categorization.27 Among them is the benefits program administered by the Board of Veterans’ Appeals (BVA).28

This Report is mainly concerned with agency review of an adjudication decision based largely—even if not always exclusively—on the evidentiary record in the proceeding from which the decision arose, whether or not the agency review is subject to yet additional layers of internal review. Hence it covers what might be called conventional appellate review. But the Report is also concerned with some adjudication decisions arising from de novo evidentiary proceedings—most, but not all, following a decision by a frontline adjudicator in an informal proceeding—that may have some appellate characteristics. Again, the PTAB is illustrative. Any more precision in defining “appellate review” is neither possible nor necessary to this undertaking. We trust that it will be clear enough to agencies whether the findings and recommendations offered here are relevant to their programs.

23 See Fed. R. App. P. 10. Fact-finding by the federal courts of appeals is limited. See, e.g.

Adam Zimmerman, The Appellate Class Action (unpublished manuscript). 24 See App. N (SSA); see also Office of the Chairman, Assessing the Impact of the Region I

Pilot Program (2013) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/ assessing-impact-region-i-pilot-program.

25 See App. L (PTO). 26 See Christopher J. Walker & Melissa F. Wasserman, The New World of Agency

Adjudication, 107 CALIF. L. REV. 141, 157–60, 178–80, 194–96 (2019); App. L (PTO); ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 163–71.

27 Compare the Civilian Board of Contract Appeals, discussed in ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 115.

28 See App. G (VA).

7

One final necessary point is that what we include under the rubric of appellate review does not depend on whether the decision under review in the context of an adjudication is subject to the APA’s formal hearing provisions.29 Most agency adjudications are not subject to those provisions, and yet many of them are conducted with sufficient procedural formality (sometimes more formality than APA adjudications) so as to demand appellate review structures and procedures the same or similar to those used in APA adjudications.30

B. Structures of Agency Appellate Review 1. Categories. Agency appellate review can be structured in a myriad of

ways.31 In his 1983 study for ACUS, Professor Cass used a typology under which he identified twenty-two possible appellate structures, fourteen of which were then in use in APA adjudications alone.32 He identified which structure each APA adjudication program employed, either by statutory requirement or administrative discretion.33 Such a cataloguing exercise that covered both APA and non-APA adjudications would be near impossible.

Most agency appellate programs, though, can comfortably fit within one of the four broad categories that follow, allowing for minor qualifications. As our

29 See 5 U.S.C. §§ 554, 556–557. 30 In reliance on the work of Professor Michael Asimow, recent ACUS recommendations

divide adjudications into three categories: (1) those subject to formal hearing provisions of the APA (or APA adjudications); (2) those in which a statute, regulation, or executive order requires an evidentiary hearing, but are not subject to those provisions; and (3) those in which no evidentiary hearing is legally required. See, e.g., Admin. Conf. of the U.S., Recommendation 2016-4, Evidentiary Hearings Not Required by the Administrative Procedure Act, 81 Fed. Reg. 94,314 (Dec. 23, 2016). Professor Asimow’s report underlying Recommendation 2016-4 calls these adjudications Type A, Type B, and Type C, respectively. See Michael Asimow, Evidentiary Hearings Outside the Administrative Procedure Act (2016) (report to the Admin. Conf. of the U.S.) [hereinafter Asimow, Evidentiary Hearings], https://www.acus.gov/report/evidentiary-hearings-outside-administrative-procedure-act-final-report; see also Michael Asimow, Best Practices for Evidentiary Hearings Outside the Administrative Procedure Act, 26 GEO. MASON L. REV. 923 (2019). For an expanded discussion, see ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, 15–24. Most of the appellate programs with which we are concerned review decisions arising from Type A and Type B adjudication—but not all. Some programs that review Type C decisions are functionally and structurally similar enough to justify coverage in this Report.

31 See Recommendation 2016-4, supra note 30, at 94,315; ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 84; see also KENNETH CULP DAVIS, DISCRETIONARY JUSTICE: A PRELIMINARY INQUIRY 145 (1969) (emphasizing “divers[ity]” of appellate agencies).

32 See Cass, supra note 2, at 157–59; cf. Weaver, supra note 2, at 253–270 (placing “review structures” into one of four categories).

33 See Cass, supra note 2, at 201–17. A summary table appears on page 217.

8

classification system reflects, an important feature of appellate review programs is whether they provide for some form of agency-head review.34

a. Direct Review of Hearing-Level Adjudicator’s Decision by Agency Head. The agency head directly reviews the decision of a hearing-level adjudicator. No agency-head delegate or intermediate review body is interposed between the adjudicator and the agency head. Review is usually either as of right at the request of a party (mandatory review) or at the discretion of the agency head (either on the application of a party or, less commonly, its own initiative). Even when review is discretionary, a statute or, more commonly, a procedural rule often sets forth a standard or criteria to guide or circumscribe the discretion.35 Many agencies that fall into this first category are traditional independent regulatory boards and commissions. They include the NLRB (review as of right in unfair-labor-practice cases), the SEC (review at the discretion of the agency head), and the MSPB (review at the discretion of the agency head).36

b. Review by an Agency-Head Delegate—Often a Multi-Member Appellate Body—Resulting in a Final Decision Unreviewable by the Agency Head. The agency head delegates her authority to review the lower-level adjudicator’s decision to a subordinate official(s), often a multi-member appellate body, who issues a final decision on the agency head’s behalf. (The agency could be an administration within an agency.37) The delegation is sometimes, though not usually, made with explicit statutory authorization. The delegation is almost always on a standing, class-wide basis rather than a one-time, individual-case basis. Review can be as of right or discretionary. Notable appellate bodies in this category include the SSA Appeals Council, and, subject to narrow exceptions and with allowances for minor qualifications, the Environmental Appeals Board (EAB) at the Environmental Protection Agency (EPA).38

c. Review by a Statutorily Authorized Appellate Decisionmaker—Often a Multi-Member Body—Resulting in a Final Decision Unreviewable by the Agency Head. Unlike in the previous category, in which the agency creates an appellate body, this category involves a statutorily created appellate body. A statute vests authority in an agency official(s) (often a multi-member board) to issue final

34 See Walker & Wasserman, supra note 26, at 174–88; Harold J. Krent, Presidential Control

of Adjudication within the Executive Branch, 65 CASE W. RES. L. REV. 1083 (2015); Weaver, supra note 2.

35 See, e.g., App. J (MSPB). 36 See App. K (NLRB); App. M (SEC); App. J (MSPB); see also App. I (EEOC). 37 See Weaver, supra note 2, at 265; see also 5 U.S.C. § 551(1) (referring to agencies “within”

agencies). 38 See App. N (SSA); App. H (EPA). While EPA has not provided for appellate review of most

EAB decisions by its Administrator, its regulations do reserve the Administrator’s authority to modify an EAB decision on her own initiative. That authority has not been exercised. EPA rules also authorize the General Counsel to provide a binding interpretation of a statute or rule in EAB proceedings. See App. H. (EPA).

9

decisions. Review by the official can be mandatory or discretionary, but is usually mandatory. The decision is unreviewable by the agency head, as a matter of either law or practice. A notable example is the VA’s Board of Veterans’ Appeals.39 Another is the Department of Agriculture’s Judicial Officer.40 The PTAB, noted above, can also be put in this category.

d. Review by an Intermediate Appellate Adjudicator—Often a Multi-Member Appellate Body—Resulting in a Decision Subject to Agency Head Review. A statute or agency rule assigns appellate review to an intermediate adjudicator, often a multi-member appellate body. (If the assignment is made by a rule, the review will be by an agency-head delegate.) That adjudicator’s decision will be subject to (usually discretionary) review by the agency head. The most notable example is DOJ’s immigration adjudication program, under which the Board of Immigration Appeals’ decisions are subject to discretionary review by the Attorney General.41 Another notable example is Department of Labor’s (DOL’s) Administrative Review Board, which decisions the Secretary of Labor recently made subject to discretionary secretarial review.42

2. Governing Law. The structure of an appellate program can be the result of statutory law or an agency rule, or a combination of both. Most agencies’ programs are administratively established by rule as far as their particulars are concerned, but there are notable exceptions, among them the PTO (for patent cases) and the Veterans Administration (for veterans benefit cases). Whatever the legal basis of an appellate program’s structure, it usually enjoys considerable administrative discretion to make the procedural choices reflected in Parts III (Findings) and IV (Recommendations) of this Report.

C. Models and Objectives of Agency Appellate Review “Modern administrative law,” as Professor Thomas Merrill chronicled nearly

a decade ago, “is built on the appellate review model of the relationship between reviewing courts and agencies.”43 This appellate review model resembles judicial review of trial-court civil litigation. In particular, the appellate court reviews a closed record, and the standard of review varies based on the comparative competence of the trial and appellate court. At one extreme, appellate courts defer to trial courts as to factual findings. At the other extreme of purely legal questions, appellate courts engage in more searching review.

39 See App. G (VA). 40 See ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 109. 41 See App. E (DOJ). See generally Walker & Wasserman, supra note 26, at 171–73. 42 U.S. Dep’t of Labor, Secretary’s Order No. 01-2020, Delegation of Authority and

Assignment of Responsibility to the Administrative Review Board, 85 Fed. Reg. 13,186 (Mar. 6, 2020).

43 Thomas W. Merrill, Article III, Agency Adjudication, and the Origins of the Appellate Review Model of Administrative Law, 111 COLUM. L. REV. 939, 940 (2011).

10

To be sure, an appellate review model is not the only one that could have developed in the administrative law context. Indeed, as Professor Merrill detailed, prior to the early twentieth century, the model was quite different, with no judicial review at all for some agency actions and mandamus-like review for others.44 The appellate review model emerged, he argued, in response to backlash over the Supreme Court’s searching review of rate and service orders of the Interstate Commerce Commission and Congress’s enactment of the Hepburn Act of 1906.45

When we think of appellate review within an agency, we may be tempted to gravitate to this familiar appellate model of judicial review of administrative action. Put differently, we may view the objectives of agency appellate review as the same as those for judicial review. That would be a mistake. Agency appellate review may well serve distinct purposes. For example, as Professor Cass outlined in his 1983 ACUS report, administrative law scholarship has often reflected two different, contrasting models for agency decision making: (1) the “judicial model,” in which “a neutral arbiter weighs evidence and ascertains facts”; and (2) the “political model,” in which “decisions do not turn on descriptive facts but on the identity of interested parties, the intensity of their interests, and on assumptions about the impact of particular decisions on future events.”46 When it comes to agency adjudication, Professor Cass ultimately concluded that these models are synthesized: the hearing-level agency adjudicators largely embrace the judicial model, whereas the appellate review function may incorporate more of the political model.47

When it comes to agency appellate review, moreover, James Freedman in 1969 identified two distinct models: (1) the “judicial model,” which consists of error correction and the application of law to fact without engaging in policymaking; and (2) the “administrative model,” which includes “deciding appeals in adjudication cases as well as formulating policy.”48 In assessing whether Congress should enact legislation to authorize intermediate appellate review boards at federal agencies, Professor Freedman emphasized that Congress should not mandate a one-size-fits-all model and set of procedures. After all, “agencies that establish Review Boards will not do so for the same reasons,” and the models they adopt “will reflect different estimates of different needs in different agencies.”49

44 See id. at 946–53. 45 See id. 953–65; see also id. at 965–972 (detailing the “entrenchment of the appellate review

model” by 1930). 46 Cass, supra note 2, at 117. 47 See id. at 121–24. One of us has made a similar observation with respect to the

constitutional tensions in agency adjudication between decisional independence and political accountability. See Christopher J. Walker, Constitutional Tensions in Agency Adjudication, 104 IOWA L. REV. 2679 (2019).

48 Freedman, supra note 3, at 558–59. 49 Id. at 575.

11

In surveying agency appellate programs and interviewing their leaders across the federal regulatory state, it becomes immediately apparent that the judicial model of appellate review is not the predominant one within agencies, although it does play an important role in many programs. Neither is there—nor should there be—one unifying model. Agencies operate appellate review programs for distinct purposes. Sometimes the agency does so because Congress has commanded a certain objective or multiple objectives in appellate review. Other times, the agency itself has exercised its discretion to focus appellate review on certain objectives, which hopefully advance the statutory purposes of the agency adjudication program as a whole.

This is not merely a point of academic interest. The purpose of appellate review will dictate—or should dictate if the agency is functioning competently—the key procedural choices an agency must make. One helpful example is whether (when the agency has a choice) appellate review is discretionary or mandatory, and, if discretionary, what criteria is used to decide which appeals to hear. An agency cannot answer those questions without considering the agency’s objectives in appellate review. Consider, for instance, the U.S. Supreme Court’s approach to managing its discretionary certiorari docket. Supreme Court Rule 10 outlines the considerations for granting review. It notes that review is “granted only for compelling reasons,” and then it sets forth three reasons that are “neither controlling nor fully measuring the Court’s discretion.” The three reasons proffered (a circuit-court conflict, a state-supreme-court conflict, and “an important question” of federal law) all underscore that the Court’s discretionary docket is not about just error correction. Indeed, Rule 10’s final sentence makes that plain, providing that a petition “is rarely granted when the asserted error consists of erroneous factual findings or the misapplication of a properly stated rule of law.”

With the vast and varied administrative adjudications taking place in the federal administrative state today, it should surprise no one that there is not one unified model of agency appellate review. Nor should there be, in light of the diverse purposes in each administrative adjudication program. What is surprising, however, is that the literature has not seemed to develop a taxonomy of objectives that would motivate agency appellate review models. At least seven such objectives emerged from our study of appellate programs. Each one merits some elaboration here.

1. Error Correction. A conventional rationale for appellate review is to correct errors in the hearing-level adjudication.50 These errors could involve articulating the legal standard, applying the law to particular facts, and perhaps even factfinding. The correction of an error itself may be a compelling justification for appellate review. Moreover, as Steven Shavell has explained, an efficient appellate review process “may render society’s investment in the appeals process economical by comparison to its improving the accuracy of the

50 See generally Chad M. Oldfather, Error Correction, 85 IND. L.J. 49 (2010).

12

trial process—an approach that, by its nature, would require extra expenditure in every case.”51 In other words, an agency may decide that, if the goal is to minimize erroneous adjudication outcomes generally, investing in appellate review is more cost-effective than investing in further improving the hearing-level agency adjudications.

A related purpose that we group with error correction is “error prevention: inducing trial court judges to make fewer errors because of their fear of reversal.”52 The threat of further review may well encourage the hearing-level adjudicators to avoid errors in the first place.

2. Inter-Decisional Consistency. Much of the agency adjudication scholarship in recent years has focused on the importance of inter-decisional consistency among hearing-level adjudications. As David Hausman put it in the immigration adjudication context, “Uniformity is both a goal of appeals processes and an indication that they are functioning properly.”53 In other words, agency appellate review should seek to ensure that like individuals are treated the same, regardless of the identity of the hearing-level adjudicator. In the patent adjudication context, Professors Michael Frakes and Melissa Wasserman have argued that inter-decisional inconsistency can be evidence of hearing-level adjudicators committing routine errors and can erode public confidence that the adjudication system is fair and nonarbitrary.54 Some scholars of the social security adjudication system have similarly argued that consistency can serve as a proxy for accuracy, apart from whatever due-process norms the former serves.55

This focus on consistency is not new to the literature. Back in the 1980s, it was captured in Jerry Mashaw’s “bureaucratic rationality” model for administrative action, which aims “to minimize the sum of error costs and administrative costs.”56 At bottom, Professor Mashaw’s concept of “bureaucratic rationality,” as Robert Kagan has summarized it, is a model of administrative adjudication that facilitates “[g]reater control and consistency” by placing the

51 Steven Shavell, The Appeals Process as a Means of Error Correction, 24 J. LEGAL STUD. 379, 381–82 (1995).

52 Id. at 425–26. 53 David Hausman, The Failure of Immigration Appeals, 164 U. PA. L. REV. 1177, 1181

(2016). 54 Michael D. Frakes & Melissa F. Wasserman, Patent Trial and Appeal Board's

Consistency-Enhancing Function, 104 IOWA L. REV. 2417, 2421–23 (2019). 55 See, e.g., Richard J. Pierce, Jr., Political Control Versus Impermissible Bias in Agency

Decisionmaking: Lessons from Chevron and Mistretta, 58 U. CHI. L. REV. 481, 510 (1990). 56 Jerry L. Mashaw, Conflict and Compromise Among Models of Administrative Justice, 1981

DUKE L.J. 181, 185 (1981). For purposes of this Report, we bracket Mashaw’s competing “professional treatment” and “moral judgement” models, which emphasize client satisfaction and fair resolution of controversies, respectively. See, e.g., id. at 190 tbl.A. These may well be additional objectives for administrative action. Based on our study, however, they do not seem to have much traction today in the rationales for agency appellate programs.

13

“overriding value” on “accurate, efficient and consistent implementation of centrally-formulated policies.”57

3. Policymaking. In the judicial appellate review context, the “creation and refinement of law” is often proffered as a core rationale.58 In the agency appellate review context, the somewhat analogous function of policymaking may be a legitimate objective. In its seminal decision in SEC v. Chenery, the Supreme Court held that an agency has the discretion to proceed by rulemaking or case-by-case adjudication in order to establish agency policy.59 There is a long history of agency heads—as well as intermediate review bodies—utilizing appellate review to establish or further develop policy for the agency as a whole.60

At first blush, it may seem like policymaking through agency appellate review falls exclusively within Professor Cass’s “political model” for administrative action.61 On closer examination, however, it is not just about implementing the agency’s (or presidential administration’s) policy preferences within the agency’s statutory authority. Effective policymaking via appellate review can also advance the “judicial model” and related concerns about error correction and inter-decisional consistency. As Judge Henry Friendly observed long ago, a critical problem with administrative action “is the failure to develop standards sufficiently definite to permit decisions to be fairly predictable and the reasons for them to be understood.”62 Policymaking via agency appellate review can lead to greater consistency in adjudication outcomes at both the hearing and appellate levels.

57 Robert A. Kagan, Inside Administrative Law, 84 COLUM. L. REV. 816, 820 (1984). See

generally JERRY L. MASHAW, BUREAUCRATIC JUSTICE: MANAGING SOCIAL SECURITY DISABILITY CLAIMS 25–26 (1983); Jerry L. Mashaw, The Management Side of Due Process, 59 CORNELL L. REV. 772 (1974).

58 Oldfather, supra note 50, at 49 (“Most depictions of appellate courts suggest that they serve two core functions: the creation and refinement of law and the correction of error.”).

59 Sec. & Exch. Comm’n v. Chenery Corp., 332 U.S. 194, 203 (1947) (“There is thus a very definite place for the case-by-case evolution of statutory standards. And the choice made between proceeding by general rule or by individual, ad hoc litigation is one that lies primarily in the informed discretion of the administrative agency.”).

60 One of us has argued that the Executive Branch in the immigration context should flip the policymaking default from agency adjudication to rulemaking. See Shoba Sivaprasad Wadhia & Christopher J. Walker, The Case Against Chevron Deference in Immigration Adjudication, 70 DUKE L.J. (forthcoming 2021), https://ssrn.com/abstract=3662827.

61 There is ample scholarship supportive and critical of agency policymaking by adjudication. See, e.g., M. Elizabeth Magill, Agency Choice of Policymaking Form, 71 U. CHI. L. REV. 1383, 1396–97 (2004); David L. Shapiro, The Choice of Rulemaking or Adjudication in the Development of Administrative Policy, 78 HARV. L. REV. 921, 922 (1965).

62 HENRY FRIENDLY, THE FEDERAL ADMINISTRATIVE AGENCIES: THE NEED FOR BETTER DEFINITION OF STANDARDS 5–6 (1962).

14

4. Political Accountability. Professor Ronald Levin has observed that the “standard federal model” for administrative adjudication continues to vest final decision-making authority in the agency head.63 One reason for that is political accountability. As noted in the Introduction, the Supreme Court has become increasingly interested as a constitutional matter in ensuring that the President has sufficient control over the bureaucracy.

In Lucia v. SEC, for instance, the Court held that ALJs at the SEC are at least “inferior officers” under the Appointments Clause, and thus must be appointed by the agency head (or the President).64 In a concurring opinion joined by Justice Gorsuch, Justice Thomas grounded his approach in the importance of political accountability in the administrative state: “[B]y specifying only a limited number of actors who can appoint inferior officers without Senate confirmation, the Appointments Clause maintains clear lines of accountability—encouraging good appointments and giving the public someone to blame for bad ones.”65 Agency appellate review, when designed to facilitate agency-head oversight, can similarly help advance this political accountability value by ensuring that the Senate-confirmed agency head, or her designee, can exercise final decision-making authority in agency adjudication.

5. Management of the Hearing-Level Adjudication Program. Due to concerns for decisional independence of the hearing-level agency adjudicators and for the separation of enforcement and adjudicative functions at the agency, the adjudicative functions hearing-level officials undertake are often quite insulated from the agency head and other high-level policymakers. Yet for political accountability and other policy reasons, and in some cases because a statute mandates a particular approach, the agency head needs to ensure that hearing-level adjudications are as fair, consistent, and efficient as possible. Some agencies may also have a related objective of preventing fraud at the hearing level. Appellate review helps the agency head promote and ensure such values in the hearing-level adjudication program. It does so through the mix of values already discussed—error correction, error prevention, inter-decisional consistency, and policymaking.

63 Ronald M. Levin, Administrative Judges and Agency Policy Development: The Koch Way,

22 WM. & MARY BILL RTS. J. 407, 412 (2013). 64 Lucia v. Sec. & Exch. Comm’n, 138 S. Ct. 2044, 2049 (2018) (citing the Appointments

Clause, U.S. CONST. art. II, § 2, cl. 2). 65 Id. at 2056 (Thomas, J., concurring); see also Jennifer L. Mascott, Who Are “Officers of the

United States”?, 70 STAN. L. REV. 443, 564 (2018) (arguing that “realignment of Article II officer status with the original meaning of the Appointments Clause would help to bring about greater democratic accountability by making it clearer that department heads are responsible at every step of the way for properly managing their agencies in the best interest of the public”).

15

6. Organizational Effectiveness and Systemic Awareness. A related benefit of appellate review is, as Professor Russell Weaver has put it, that it “helps the agency head gain greater awareness of how a regulatory system is functioning.”66 Such awareness does not just assist the agency head and agency leadership more generally in tailoring training and instruction for the agency’s adjudicators. It also helps the agency head consider whether adjustments to the regulatory scheme are necessary via rulemaking, precedential adjudication, agency guidance, legislative recourse, or other means. Such awareness is even more critical with respect to agencies that have substantial enforcement or similar regulatory responsibilities. As Melissa Wasserman and one of us have observed in the patent adjudication context:

This benefit is quite significant to the Patent Office. The Patent Office employs more than 8,000 patent examiners that process over 600,000 patent applications per year. As one of us [Wasserman] has explored in a series of papers, the quality and consistency of patent examination varies widely, and effective training can play a significant role in improving agency adjudication at the patent-issuing level. The better the Director understands how the regulatory system is functioning, the better positioned she is to address systemic issues through further guidance and training. Final decision-making authority of PTAB decisions would aid on that front.67

7. Reduction of Federal-Court Litigation. A final objective for agency appellate review programs is to reduce the number of cases that get appealed to the federal courts. This is an objective that Congress cares about generally when creating administrative agencies and procedures. One of ACUS’s five main purposes, in fact, is “to reduce unnecessary litigation in the regulatory process.”68 And this objective is particularly salient in the agency adjudication context, where adjudication is often perceived as a cheaper, more efficient alternative to traditional litigation in federal court.

Consider, for instance, Congress’s creation of the PTAB in the America Invents Act of 2011.69 The America Invents Act was enacted, in part, to respond to the growing criticism that the Patent Office issues too many invalid patents, which harms consumer welfare.70 Although federal courts can invalidate invalid patents, the cost of patent litigation had skyrocketed over the prior decade. This has hampered the ability of patent holders with meritorious claims to challenge

66 Weaver, supra note 2, at 289. 67 Walker & Wasserman, supra note 26, at 177–78 (footnotes omitted). 68 5 U.S.C. § 591(3). 69 Leahy-Smith America Invents Act, Pub. L. No. 112-29, 125 Stat. 284 (2011). 70 Michael D. Frakes & Melissa F. Wasserman, Does the U.S. Patent and Trademark Office

Grant Too Many Bad Patents?: Evidence from a Quasi-Experiment, 67 STAN. L. REV. 613, 621 (2015).

16

invalid patents in federal court. The new PTAB adjudication proceedings were designed to create a cheaper, faster alternative to district court litigation.71

Agency appellate review can play an important role in filtering out cases from reaching federal courts—thus reducing the use of federal court resources for administrative actions and making the administrative process cheaper and more accessible for everyday Americans.

D. Conference Recommendations and Model Appellate Rules 1. Recommendations 68-6 and 83-3. ACUS has issued two cross-agency

recommendations on appellate review in APA adjudications: Recommendation 68-6, Delegation of Final Decisional Authority Subject to Discretionary Review by the Agency,72 and Recommendation 83-3, Agency Structures for Review of Decisions of Presiding Officers Under the Administrative Procedure Act.73 Both favor discretionary rather than mandatory agency-head review. And both provide for, in many cases at least, an intermediate level of review: in the case of 68-6, by an “intermediate appellate body”; in the case of 83-3, by such a body or individual delegate. Recommendation 83-3 is perhaps best seen as an elaboration of Recommendation 68-6. A few differences are noted below.

Recommendation 68-6 asked agencies with “substantial caseloads” to consider establishing—and Congress to amend the APA “as necessary to clarify” their authority to establish74—either (1) “intermediate appellate boards” or (2) procedures for making presiding officers’ (usually ALJs’) decisions final, subject to summary affirmance or discretionary review by the agency head. Appellate boards, Recommendation 68-6 provided, should be authorized to make final

71 See Walker & Wasserman, supra note 26, at 157–58. 72 Recommendation 68-6, supra note 3. No ACUS recommendation specifically discusses

appellate review in non-APA adjudication programs except for Recommendation 2016-4, supra note 30, which provides that agencies should provide for such review, but does not elaborate on what form it should take. Professor Asimow’s ACUS-sponsored sourcebook on adjudication addresses, and makes some recommendations concerning, certain features of appellate review in non-APA adjudications. See ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 84–85. Our recommendations are in accord.

73 Recommendation 83-3, supra note 2. A related recommendation on which we draw is Admin. Conf. of the U.S., Recommendation 71-1, Interlocutory Appeal Procedures, 38 Fed. Reg. 19,787 (July 23, 1973). The associated report is Ernest Gellhorn & Paul B. Larsen, Interlocutory Appeal Procedures in Administrative Hearings, 70 MICH. L. REV. 109 (1971). ACUS has issued a number of agency-specific recommendations on appellate review. See, e.g., Recommendation 2013-1, Improving Consistency in Social Security Disability Adjudication, 78 Fed. Reg. 41,352 (July 10, 2013); Recommendation 90-4, Social Security Disability Program Appeals Process: Supplemental Recommendation, 55 Fed. Reg. 34,213 (Aug. 22, 1990); Recommendation 87-7, A New Role for the Social Security Appeals Council, 52 Fed. Reg. 49,143 (Dec. 30, 1987); Recommendation 86-5, Medicare Appeals, 51 Fed. Reg. 46,987 (Dec. 30, 1986). For the associated reports, see Adjudication, ACUS, https://www.acus.gov/adjudication#appellate_systems (last visited Nov. 24, 2020).

74 See Cass, supra note 2, at 148 (noting ambiguity, reflected in Recommendation 68-3, as to whether 5 U.S.C. § 557 requires agency-head review).

17

decisions, subject to discretionary review by the agency head, whether on a party’ motion or its own initiative. When direct agency-head review of ALJ decisions is provided, the agency should be able to “accord” finality to the decision either by declining review or summarily affirming it unless a party makes a “reasonable showing that” a “prejudicial error was committed” or the “initial decision embodies” a “finding . . . of material fact which is erroneous or clearly erroneous” or an erroneous legal conclusion. The agency head can also review decisions that present an “important” issue of “law or policy.”

Recommendation 83-3 addressed the “organizational structures which agencies establish to review” decisions of presiding officers. ACUS concluded that existing appellate programs were too varied to prescribe any “single structure” and that, “[b]y and large,” the existing “structures seem[ed] well adapted to the particular circumstances of the agency.” Instead, ACUS provided “general guidance” to agencies in establishing new and in reviewing existing structures. Its most important recommendations were as follows:

a. Agency Head Review (§ 1). Congress should not require agency heads to review every hearing-level decision. Review should be discretionary. Congress should authorize agency heads to delegate—on an “ad hoc” or class-wide basis—review authority to “subordinate” officials “either with the possibility for further review by the agency head in his discretion or without further administrative review.”

b. Forms of Delegation (§ 2). Agency heads that have delegable review authority should delegate it on a class-wide (not ad hoc) basis in any system that adjudicates a “substantial number of cases.” If the agency head retains further review authority, it should “be exercisable” only when “important policy issues are presented” or the “delegate erroneously interpreted agency policy.” “Multilevel review of purely factual issues should be avoided.”

c. Agency Head Choice of Delegate (§ 3). Delegation can be to an individual subordinate, a multi-member board, or, in the case of a multi-member agency (e.g., SEC), a panel of the agency’s members. Individual delegates are favored when the cases are voluminous, “simple,” and mostly fact-dependent. Board delegates are favored when the case volume is still “substantial,” but the cases are “more complex” and time-consuming. Agency panels, in the case of multi-member agencies, are favored when the caseload is large, and it is “difficult” to “elaborate or clarify” agency policy (especially through rulemaking) in “a manner that will substantially limit the number of significant policy issues presented in adjudications.”

d. Standards for Granting Review (§ 4). Agencies need not necessarily provide review as of right before the delegate. They may wish to authorize the delegate to decline review in “routine cases” absent a “reasonable showing” that the ALJ committed “prejudicial procedural error” or that the ALJ’s decision “embodies” a “finding or conclusion of material fact which is erroneous or clearly erroneous,” a “legal conclusion which is erroneous,” or “an exercise of discretion or decision of law or policy which is important and should be reviewed.” Sua

18

sponte review should be provided only for “policy issues.” This standard tracks the standard of Recommendation 68-6 addressing when agency heads should review decisions of ALJs.

Small differences—perhaps unintended—between Recommendation 68-6 and Recommendation 83-3 will be noticed. One is that Recommendation 68-6 seems to contemplate that intermediate appellate review should be as of right (although it does provide that agencies might “[r]estrict the scope of review”), whereas Recommendation 83-3 says that review need not necessarily be as of right. Recommendation 83-3 provides standards governing when review might be withheld in “routine” cases. Moreover, Recommendation 68-6 contemplates intermediate review by a multi-member body, whereas Recommendation 83-3 provides that in some cases review might best be undertaken by a single adjudicator.

2. Model Appellate Rules. In 2018, a working group commissioned by ACUS’s Office of the Chairman issued a substantial revision to ACUS’s Model Adjudication Rules.75 Rules 400–450 (“Administrative Review”) offer model rules for appellate practice, both in APA and similarly formal non-APA adjudications.76 They reflect the working group’s careful review of, among other sources, numerous agency rules, the Federal Rules of Appellate Procedure, and several federal circuits’ rules. Rules 400–450 are reproduced, with the reporter’s comments, in Appendix B, and they are relied on in this Report. The Rules were not put before ACUS’s Assembly for a vote. Hence they do not have the status of a formal recommendation.

E. Judicial Review Just a few observations on this vast subject will be useful here.77 Most final

agency adjudication orders are subject to judicial review in a federal court, if not under a program-specific statute then under the APA.78 The order must be

75 MODEL RULES, supra note 1. 76 Id. The reporter was Professor Kent Barnett. One of the authors of this Report (Wiener)

chaired the subgroup assigned to draft the appellate rules. 77 For more information on judicial review of agency adjudication orders, see MICHAEL HERZ,

RICHARD MURPHY & KATHRYN WATTS, EDS., A GUIDE TO JUDICIAL AND POLITICAL REVIEW OF FEDERAL AGENCIES 1–257 (ABA, 2d ed. 2015); ABA SECTION OF ADMIN. LAW & REG. PRAC., A BLACKLETTER STATEMENT OF FEDERAL ADMINISTRATIVE LAW 31–66 (2d ed. 2013). ACUS will soon publish a sourcebook, written by Professor Jonathan Siegel, on federal statutes governing judicial review of agency action. See JONATHAN SIEGEL, ADMIN. CONF. OF THE U.S., SOURCEBOOK OF FEDERAL JUDICIAL REVIEW STATUTES (forthcoming), https://www.acus.gov/research-projects/sourcebook-federal-judicial-review-statutes (last visited Nov. 24, 2020).

78 See 5 U.S.C. §§ 701–706. Jurisdiction is conferred either by a program-specific statute or the general federal-question jurisdictional statute, 28 U.S.C. § 1331. The APA (in particular, 5 U.S.C. § 702) does not itself confer jurisdiction on federal courts. See Califano v. Sanders, 430 U.S. 99 (1977). A cause of action is provided either by a program-specific statute or, if there is none, then by the APA itself. See 5 U.S.C. § 704.

19

final.79 When it becomes final depends on the statute or regulations governing the particular program.80

Before seeking judicial review, litigants generally must have exhausted any required administrative remedies—that is, for purposes of this Report, availed themselves of any required agency appellate review. The review must specifically be required by statute or regulation. If agency appellate review is optional, then, under the APA, a litigant may seek judicial review without first seeking agency appellate review.81 Most agency rules of practice are drafted with these principles in mind.

Jurisdiction to review adjudication orders lies in a federal district court, in the first instance, unless a specific statute provides for review in an Article I court or directly in an Article III federal court of appeals. A number of statutes do just that, including those governing many adjudications under the APA’s formal hearing provisions.82 The decisions of about half the adjudication programs studied in this Report are subject to direct review in the courts of appeals.83 Venue often lies not only in the circuit where the petitioner resides or the case arose, but also in the D.C. Circuit.84 A few agencies’ decisions are subject to review only in a single circuit.85

Under the APA’s scope-of-review provision, a reviewing court must “set aside” an adjudication order that, among other things, is arbitrary and capricious, contrary to law, or, in the case of its factual findings, “unsupported by substantial evidence.”86 Program-specific review statutes sometimes specify a scope and standard of review. Most are substantially the same as the APA’s.87 If a court finds an adjudication decision to be erroneous on an issue of fact or law, it normally remands the case to the agency to revisit the matter anew

79 See 5 U.S.C. § 704. 80 See MODEL RULES, supra note 1. 81 See Darby v. Cisneros, 509 U.S. 137, 143 (1993). Darby interpreted 5 U.S.C. § 704, which

states that unless a statute provides otherwise, “agency action otherwise final is final . . . whether or not there has been presented or determined an application for a declaratory order, for any form of reconsideration, or, unless the agency otherwise requires by rule and provides that the action meanwhile is inoperative, for an appeal to superior agency authority.” See generally MODEL RULES, supra note 1, at 78 (Comment 4); John F. Duffy, Administrative Common Law in Judicial Review, 77 TEX. L. REV. 113, 158–60 (1998).

82 See Admin. Conf. of the U.S., Recommendation 75-3, The Choice of Forum for Judicial Review of Administrative Action, 40 Fed. Reg. 27,926 (July 2, 1975). One statute provides for direct review in the courts of appeals of a number of agencies’ decisions in APA adjudications. See Administrative Orders Review Act, 28 U.S.C. §§ 2341–2351.

83 See Apps. C–N. 84 See, e.g., 28 U.S.C. § 2343. 85 See, e.g., 5 U.S.C. § 7703(b)(1)(A) (specifying the Federal Circuit for review of most MSPB

orders). 86 5 U.S.C. § 706. 87 See, e.g., 5 U.S.C. § 7703(c) (MSPB).

20

rather than deciding the issue itself and directing a particular disposition before the agency. This “ordinary remand rule,” which one us has explored elsewhere, has important implications for agency appellate review.88

A final aspect of judicial review that bears mention is the issue of agency acquiescence in decisions of the courts of appeals. Agencies must, of course, follow decisions of the Supreme Court. Whether and when they must acquiesce in the decisions of particular circuits—which, on any given issue of law, may conflict—can often be a difficult question.89 Agency appellate reviewers must sometimes answer that question when adjudicating cases—often without knowing in which circuit a litigant will seek review—and they must sometimes decide what directives to give lower-level adjudicators and by what means.

II. STUDY METHODOLOGY

The landscape of agency appellate review is vast and varied, and little trans-substantive attention has been paid to it in nearly forty years. Accordingly, our study is necessarily exploratory and qualitative. To structure our study methodology, we began by reviewing the literature on agency appellate review, ACUS’s recommendations and reports on the subject,90 and the database maintained by Stanford Law School that reflects ACUS’s extensive research on administrative adjudication.91 We also sought suggestions from a number of leading academic experts on adjudication as well as various current and former agency officials.92

Based on this initial research, we settled on twelve agencies—in some cases, limiting ourselves to particular adjudication programs of special importance within them—for focused study. They are:

• Department of Health and Human Services (HHS): Departmental Appeals Board—Medicare Appeals Council;

• Department of Homeland Security (DHS): U.S. Citizen and Immigration Services (USCIS), Administrative Appeals Office;

• Department of Justice (DOJ): Executive Office of Immigration Review (EOIR)—Board of Immigration Appeals (BIA);

• Department of Labor (DOL): Administrative Review Board;

88 See Christopher J. Walker, The Ordinary Remand Rule and the Judicial Toolbox for Agency Dialogue, 82 GEO. WASH. L. REV. 1553 (2014).

89 See generally Samuel Estreicher & Richard L. Revesz, Nonacquiesence by Federal Administrative Agencies, 98 YALE L.J. 679 (1989).

90 See Part I.D supra. They include ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20. 91 See Stanford Law School, Adjudication Research, http://acus.law.stanford.edu (last visited

Nov. 25, 2020). Of particular relevance are two of the twelve tables (which the database calls “reports”) on the homepage: “Types and Hearings and Appeals” and “Ability to Appeal.”

92 See infra note **.

21

• Department of Veterans Affairs (VA): Board of Veterans Appeals (BVA);

• Environmental Protection Agency (EPA): Environmental Appeals Board (EAB);

• Equal Opportunity Employment Commission (EEOC); • Merits Systems Protection Board (MSPB); • National Labor Relations Board (NLRB); • Patent and Trademark Office (PTO): Patent Trial and Appeals

Board (PTAB); • Securities and Exchange Commission (SEC); and • Social Security Administration (SSA): Appeals Council. In compiling this list, we sought to include well-known agencies/programs

representative of the predominate structures of appellate review identified in Part II.B, both executive-department agencies (e.g., DOL) and independent agencies (e.g., NLRB, SSA), both traditional regulatory and benefits-conferring agencies, and so forth. We also considered volume. The selected agencies, for the most part, adjudicate a relatively large number of cases, and several of them (SSA, VA, and PTAB) administer among the government’s largest adjudication programs.93

For each selected agency/program, we prepared an extensive written overview that explains what it does and, in table form, its key institutional and procedural attributes as ascertainable from publicly available sources (mostly statutes, regulations, guidance documents, website explanatory materials, and the like). Appendix A provides the template we used for these overviews. The twelve appellate program overviews appear in Appendices C–N (in alphabetical order by agency name), which due to their length are posted on ACUS’s website.94 We usually cite to the overviews in this Report rather than the original sources unless the latter requires emphasis.

To better understand how these appellate programs operate in practice, we conducted semi-structured interviews with at least one high-ranking official at each agency—in most cases the head of the agency appellate program. In total, we interviewed more than two dozen agency officials. Most of these interviews

93 Our list overlaps considerably with the list of twelve agencies/programs Professor Asimow singled out for detailed review in his report underlying ACUS Recommendation 2016-4, see Asimow, Evidentiary Hearings, supra note 30, and his ACUS-commissioned sourcebook on adjudication, see ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 107–82. The best practices in his report (and incorporated into the recommendation) were based mostly on his study of the twelve agencies. See ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 6.

94 Admin. Conf. of the U.S., Agency Appellate Systems, https://www.acus.gov/research-projects/agency-appellate-systems (last visited Nov. 25.2020) (the Appendix file is located under the Project Documents heading).

22

were in person, but some were conducted by phone. To prepare for the interviews, we constructed an interview script that largely tracks the topics and structure of the findings section of the Report (Part III). Among other things, we focused on the agency’s objectives for appellate review, the challenges the appellate program faces, and any best practices the agency has adopted. To develop the interview script, we consulted with various scholars and agency officials. During the interviews, we largely followed the script, but we also asked follow-up questions both as to answers given during the interview and as to questions we had based on the extensive agency overview we had created in advance of the interview.

Specific responses, we told interviewees, would not be attributed to them or their agency. The exception would be for responses that identified what interviewees considered best practices, especially those they might commend to other agencies for consideration. Our findings (Part III) and recommendations (Part IV) are written accordingly. Most of our findings are based on the above-noted sources. Some are based on additional research of the agencies under review as well as research into other agency appellate programs.

It should go without saying that there are significant methodological limitations in our study design. Our findings and recommendations are based largely on the existing literature, publicly available sources, and interviews with high-ranking officials in each adjudication program. Publicly available sources often do not reflect the realities of administrative practice. Interviews of high-ranking agency officials can shed important light on those realities. But interviews are necessarily limited by the questions asked, by the perspectives and knowledge of the interviewees, and by other explicit and implicit factors that influence information gathering via semi-structured interviews. This is, in essence, an exploratory study. It would be improper to, among other things, attempt to generalize from our findings to other agency adjudication programs. Indeed, a core finding from our study is that agency appellate programs are not all the same.

Our hope is that the findings and recommendations we present in Part III and Part IV, respectively, will be of use for agencies that operate appellate programs and will spark further discussion and attention to agency appellate programs—in terms of research and empirical analysis as well as practical, doctrinal, and theoretical developments.

III. FINDINGS

Based on our interviews with agency officials and review of publicly available sources on the agency appellate review programs detailed in Part II, we organize our findings in seven subparts. We first explore the law of appellate review established in these programs (Part III.A), followed by the characteristics of agency decisionmakers (Part III.B) and the various criteria for selecting cases for appellate review (Part III.C). We then focus on the decision-making process

23

and types of decisions (Part III.D), the characteristics for administration and bureaucratic oversight (Part III.E), and the mechanisms for public disclosure and transparency (Part III.F). Part III concludes with some observations on the common challenges in agency appellate programs identified by the interviewees (Part III.G). Following Part III, we detail our recommendations in Part IV.

A. Law of Appellate Review 1. Codified Objectives of Appellate Review. As noted in Part I.C, the

objective(s) of an agency’s appellate review program will or should drive any number of procedural choices an agency must make—among them: whether review should be as of right or discretionary, whether and when decisions should be designated as precedential, and what the scope and standard of review should be.95 Many of the key choices are addressed in the Report.

We received thoughtful answers from most interviewees, to be sure, about the objective(s) of their programs. But we found few agency rules that actually address them, at least with any useful specificity. One program’s rules of practice that do so forthrightly is the BIA’s. They identify as one purpose of review the provision of “clear and uniform guidance”—not only to immigration judges, but also enforcement-level personnel at DHS and the “general public”—“on the proper interpretation and administration” of federal immigration law.96 Interviewees generally agreed on the value of this sort of codification. A few programs include brief statements of purpose in explanatory materials on their websites.97

2. Sources, Types, and Coverage of Procedural Law. Some statutes prescribe agency procedural law in some detail. But most leave key procedural choices to agencies. Those choices are embodied in a number of different types of legal pronouncements of varying degrees of formality, among their other differences. A complete taxonomy is not possible here.98 A few observations will suffice.

Nearly all of the appellate programs we studied are governed by formal procedural rules—often called rules of practice99—codified in the Code of

95 See Jerry L. Mashaw, Reforming the Bureaucracy: The Administrative Conference Technique, 26 ADMIN. L. REV. 261, 267 (1974) (arguing that rational administrative reform depends on the identification of “a good conceptual model of various types of agency functions” and a clear statement of “what a particular procedural regime [is] supposed to do”).

96 8 C.F.R. § 1003.1(d)(1). 97 See, e.g., U.S. Citizenship & Immigration Servs., Administrative Appeals,

https://www.uscis.gov/administrative-appeals (last visited Nov. 25, 2020) (providing that appeals office conducts review “in order to promote consistency and accuracy in the interpretation of immigration law and policy”).

98 For a list of the most common authorities, see Admin. Conf. of the U.S., Recommendation 2018-5, Public Availability of Adjudication Rules, 84 Fed. Reg. 2142 (Feb. 6, 2019).

99 See, e.g., App. K (NLRB); App. M (SEC).

24

Federal Regulations (CFR).100 Appellate rules are often integrated into broader procedural rules governing both hearings and appeals—a desirable arrangement whenever possible—and some procedural rules are combined with substantive rules.101 These appellate rules serve much the same function as, and observe forms similar to, the Federal Rules of Appellate Procedure.

CFR-codified rules of practice, though procedural and hence exempted from APA notice-and-comment requirements, are best classified as legislative rules under now-prevailing usage.102 They are legally binding on agency and litigants alike. That means they are externally enforceable, which is to say enforceable on judicial review of agency action.103 One recent ACUS Recommendation calls them “agency-promulgated rules of procedure with legal effect.”104 At some agencies, case-specific or standing procedural orders may supplement these rules.

Any number of other, less-formal agency pronouncements—which some might call internal administrative law—govern or explain appellate procedures and processes, as they do hearing-level procedures and processes.105 They include (to use commonly used names) manuals, policy statements, guidance, guides, outlines, forms, website explainers, and internal memoranda. Some are publicly available, others not. What legal classification they enjoy and what legal effect they possess have often confounded courts and commentators. No doubt many qualify as policy statements—often referred to, along with interpretive rules, as guidance documents—under the APA.106 Some may be binding, but only internally on adjudicators and other agency staff through internal disciplinary mechanisms, but otherwise lack the force of law. Others, though, may be binding against the agency—that is, externally enforceable

100 An exception is DOL’s Administrative Review Board. See App. F (DOL). The Board applies certain specified provisions of the Federal Rules of Appellate Procedure “to the extent practicable.” DEP’T OF LABOR, RULES OF PRAC. & PROC. BEFORE THE ADMIN. REV. BOARD, https://www.dol.gov/agencies/arb/resources/rules (last visited Nov. 25, 2020). The Board also has several CFR-codified rules of practice for cases dealing with certain specific subject matters. See id. (providing non-exhaustive list with links).

101 See, e.g., Sec. & Exch. Comm’n, 17 C.F.R. subpt. D. 102 See Admin. Conf. of the U.S., Recommendation 2019-6, Independent Research by Agency

Adjudicators in the Internet Age, 84 Fed. Reg. 71,350, 71,351 & n.18 (Dec. 27, 2019); see also, e.g., Cary Coglianese, Public Availability of Agency Guidance Documents (2019) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/consultant-report-public-availability-agency-guidance-documents; Ronald M. Levin, Rulemaking and the Guidance Exception, 70 ADMIN L. REV. 263 (2018).

103 See, e.g., Gillian E. Metzger & Kevin M. Stack, Internal Administrative Law, 115 MICH. L. REV. 1239, 1284–86 (2017).

104 Admin. Conf. of the U.S., Recommendation 2018-5, Public Availability of Adjudication Rules, 84 Fed. Reg. 2142 (Feb. 6, 2019) (preamble at 1).

105 See, e.g., Metzger & Stack, supra note 103, at 1284–86; see also Peter L. Strauss, The Rulemaking Continuum, 41 DUKE L.J. 1463, 1477 (1992).

106 See 5 U.S.C. §§ 551(4), 553(b).

25

through judicial review—on one legal basis or another, even if the agency intended otherwise.107

A notable example is the SSA Appeals Council’s Hearings, Appeals and Litigation Manual (HALLEX),108 which “conveys” “procedural guidance” for adjudicative staff and “defines procedures for carrying out policy and provides guidance for . . . adjudicating claims” at both the hearing and Appeals Council level. SSA deems HALLEX to be an internally enforceable policy statement—enforceable through discipline and, arguably, as a basis for the Appeals Council’s assignment of error to ALJ decisions.109 Circuit courts have split on whether it is binding against SSA on judicial review. If it is binding, non-compliance can be the basis for the assignment of error, and thus reversal and remand by the federal court.110

A third and sometimes overlooked source of procedural law relevant here are agency decisions. Agency decisional law sometimes addresses such important matters as the standard of review and the interpretation of CFR-codified rules of practice. It may also address the procedural and evidentiary law governing hearings. Decisional law is generally both internally and externally (i.e., judicially) enforceable unless, under agency rule, it is deemed nonprecedential, in which case it governs only the particular case in which it arose.

We found nearly all of the CFR-codified rules of practice we reviewed to be at least serviceable. Many compare favorably, in both style and substance, to the Federal Rules of Appellate Procedure, especially if allowance is made for the superior resources enjoyed by the federal judiciary’s rulewriters.

Some agencies’ rules of practice are more comprehensive than others.111 Not all agencies’ rules address such important procedural matters as the standard of review, the content of the record on appeal, aggregation, issue exhaustion,

107 See, e.g., Metzger & Stack, supra note 103, at 1278, 1284–86. 108 SOC. SEC. ADMIN., HEARINGS, APPEALS, AND LITIGATION LAW MANUAL § I-1-0-1

[hereinafter HALLEX], https://www.ssa.gov/OP_Home/hallex/I-01/I-1-0-1.html (last visited Nov. 25, 2020). Cf. U.S. Citizenship & Immigration Servs., AAO Practice Manual—About the Administrative Appeals Office Practice Manual, https://www.uscis.gov/administrative-appeals/aao-practice-manual (last visited Nov. 26, 2020).

109 HALLEX, supra note 108, § I-3-3-3 (“While sub-regulatory issuances (including the Hearings, Appeals and Litigation Law manual) are not categorized as ‘law’ for this purpose, the Appeals Council will grant review under the ‘error of law’ standard if an administrative law judge (ALJ) does not adhere to sub-regulatory policy or procedure and the ALJ’s non-compliance results in a due process violation.”). But see Envtl. Prot. Agency, Streamlining Procedures for Permit Appeals, 85 Fed. Reg. 51,650, 51652 (Aug. 21, 2020) (eliminating rule so as to clarify that EAB’s scope of review “does not extend to the [EPA’s] compliance with discretionary policies” and thereby make EAB’s review “more akin to that of the federal courts”).

110 Timothy H. Gray, Manual Override? Accardi, Skidmore, and the Legal Effect of the Social Security Administration’s HALLEX Manual, 114 COLUM. L. REV. 949 (2014).

111 Among especially comprehensive (and otherwise well drafted) rules are the SEC’s, see 17 C.F.R. §§ 201.400–201.490, and the MSPB’s, see 5 C.F.R. pt. 1201.

26

administrative exhaustion, the submission of evidence not part of the record below, the assignment of cases to panels, interlocutory appeals, amicus participation, the availability of and criteria for oral argument, reconsideration, and the precedential status of decisions. (Not all of these subjects will be relevant to every adjudication program.) These and other important matters are sometimes addressed, if they are addressed at all, only in guidance documents or adjudication decisions. The choice of placement sometimes seems haphazard.

Addressing all significant procedural matters in a CFR-codified legislative112 rule of practice obviously has numerous advantages. Doing so should make them more readily accessible to litigants and their lawyers (especially lawyers who do not regularly practice before the agency), agency staff, courts, reviewing courts, and policymakers.113 It also serves the rule-of-law function of committing the agency to follow standardized procedures in similarly situated cases.114 Where the agency wishes to reserve discretion with respect to a particular procedural matter, it can do so in the rule. Legislative rules often do just that, as when they commit to the adjudicator discretion as to whether to grant review, hear oral arguments, or permit amicus briefs. Third, doing so may avoid litigation disputes as to whether a particular pronouncement is enforceable against the agency on judicial review.115 Finally, it removes sticky questions that sometimes arise as to whether and when certain adjudicators must follow non-legislative (internal agency) rules and, if they do not, what consequences follow.116

No hard-and-fast rule dictates which aspects of appellate procedure should be addressed where. Certainly, some minor procedural matters—e.g., brief formatting rules, citation requirements, and filing instructions—can or even should be left to instructional documents like practice manuals that appear on the agency’s website. But there are certain significant matters that all agencies should consider including in their CFR-codified rules of practice and that sometimes escape inclusion there. Sections 400–500 of the Model Appellate Rules (Appendix B) provide helpful guidance as to what procedural matters agencies should consider addressing. Matters addressed there, if applicable to a program, should generally be addressed in a rule of practice rather than relegated to a non-legislative rule unless the agency has a good reason for doing otherwise. Other guidance as to what matters should be addressed in rules of

112 A rule’s placement in the CFR does not itself answer the question whether it is legislative. See, e.g., Health Ins. Ass’n of Am., Inc. v. Shalala, 23 F.3d 412, 423 (D.C. Cir. 1994) (stating that “publication in the Code of Federal Regulations, or its absence” is only “a snippet of evidence of agency intent” that the published pronouncement has binding effect).

113 Cf. Emily Bremer, Reckoning with Adjudication’s Exceptionalism Norm, 69 DUKE L.J. 1749, 1783–84 (2020) (discussing inaccessibility of adjudication rules generally).

114 See generally Elizabeth Magill, Agency Self-Regulation, 77 GEO. WASH. REV. 859 (2009). 115 Sometimes a court will hold an agency to a non-legislative rule under one doctrine or

another. See Metzger & Stack, supra note 103, at 1284–86 (discussing application of Accardi doctrine to non-legislative rules); see also Gray, supra note 110 (discussing SSA’s HALLEX).

116 See, e.g., Social Sec. Admin. v. Butler, 2016 WL 4547551 (M.S.P.B. 2016).

27

practice may be found by the Federal Rules of Appellate Procedure, provisions of the United States Code governing the structure of the courts of appeals, and the various federal circuits’ rules.117 One of our recommendations identifies significant matters that should be addressed in rules of practice but do not always appear in some agencies’ rules.

Two last findings bear mention. First, some of the rules of practice we reviewed identify the statutory source of particular rules. The MSPB’s rules, among others, are especially good.118 Second, some rules of practice begin with helpful background on the appellate programs they govern: the structure and composition of the reviewing authority, the source of its powers, its place in the overall adjudication system of which it is part, and so forth.119 (Sometimes this background appears only in website explanatory materials.) Both practices have obvious benefits for private litigants, reviewing courts, and policymakers.

3. Public Input in the Development of Procedural Law. As noted above, procedural rules are exempt from the APA’s notice-and-comment requirements.120 ACUS has recommended that agencies voluntarily seek public comment on procedural rules unless the costs of doing so clearly outweigh the benefits.121 Our research has disclosed that, in recent years, agencies have usually sought, and benefitted from, public comments on significant rule changes.122 (The notice often makes clear, as it generally should, that the agency is acting voluntarily by seeking public comment.) These include, notably, significant changes to appellate-specific rules.123 The solicitation of public comment on procedural rules—long a feature of the federal-court rule-writing process124—should remain a regularly followed practice. Seldom will the cost of

117 See 28 U.S.C. §§ 41–49. 118 See 5 C.F.R. § 1201. 119 See, e.g., id.; 40 C.F.R. § 22.4; Bd. of Veterans’ Appeals, 38 C.F.R. § 20.100–101. 120 See 5 U.S.C. § 553(b). 121 Admin. Conf. of the U.S., Recommendation 92-1, The Procedural and Practice Rule

Exemption from the APA Notice-and-Comment Rulemaking Requirements, 57 Fed. Reg. 30,102 (July 8, 1992).

122 See, e.g., Envtl. Prot. Agency, Streamlining Procedures for Permit Appeals, 85 Fed. Reg. 51,650 (proposed Dec. 3, 2019) (to be codified at 40 C.F.R. pts. 1, 49, 71, & 124); Sec. & Exch. Comm’n, Amendments to the Commission’s Rules of Practice, 81 Fed. Reg. 50,212 (July 29, 2016) (to be codified at 17 C.F.R. pt. 201).

123 See, e.g., U.S. Dep’t of Labor, Secretary’s Order No. 01-2020, Delegation of Authority and Assignment of Responsibility to the Administrative Review Board, 85 Fed. Reg. 13,186 (Mar. 6, 2020).

124 See U.S. Courts, How the Rulemaking Process Works, https://www.uscourts.gov/rules-policies/about-rulemaking-process/how-rulemaking-process-works (last visited Nov. 25, 2020). The Judicial Conference affirmatively solicits suggestions for rule changes on its website. U.S. Courts, How to Suggest a Change to Federal Court Rules and Forms, https://www.uscourts.gov/ rules-policies/about-rulemaking-process/how-suggest-change-federal-court-rules-and-forms (last visited Nov. 25, 2020). Agencies might consider doing likewise.

28

soliciting public comment exceed the benefits of doing so. Of course, agencies will sometimes be justified in making minor and technical changes to procedural rules without soliciting public comment.

B. Decisionmakers Most agency appellate adjudicators fall into one of two broad categories:

Senate-confirmed, presidentially appointed agency heads, including the members of boards and commissions; and agency officials appointed by agency heads, usually to appellate boards and councils. (Before Lucia v. SEC,125 the latter were sometimes appointed by other agency officials.)

Agency-head adjudicators enjoy no fixed statutory terms, as in the case of the heads of cabinet departments, or relatively short, fixed terms (five years is common), as in the case of board and commission members. Many board and commission members and a few other agency-head adjudicators enjoy for-cause protection. (Whether such protection is constitutional is another matter; the Supreme Court has been grappling with the question.126) Some boards and commissions, by statutory mandate, are subject to party-balancing requirements. Few statutes prescribe qualification requirements.

Adjudicators appointed by agency heads have either no terms or comparatively long terms (e.g., twelve years at the EAB), though there are exceptions. Some enjoy for-cause removal protection, if only under the generally applicable civil service laws enforced by the MSPB rather than adjudicator-specific rules.127 They are often thought of (though the term is imprecise) as “career officials.” Their legal appointment and compensation status vary considerably. Some, for instance, are hired as lawyers in the excepted service and paid on the GS scale; some are members of the Senior Executive Service. There are other possibilities.128 Many are selected from among existing agency employees (say, the ranks of hearing-level adjudicators or appellate-level staff attorneys); some are selected from among applicants (internal and external) recruited through publicized job-opening notices. Qualifications are usually prescribed, whether by statute, agency rule, ad hoc agency policy, or otherwise. Most would likely be considered merit, as opposed to political, appointees, although those terms are slippery.

A few indicia of impartiality are important. Most agency-head adjudicators are not governed by impartiality statutes or rules. Their impartiality obligations arise only as a matter of constitutional due process. Adjudicators appointed by

125 138 S. Ct. 2044 (2018). 126 See, e.g., Seila Law LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183 (2020). 127 Constitutional questions have arisen as to whether such adjudicators who issue final,

unreviewable agency decisions are principal officers, and, if so, what consequences follow. See supra notes 12–15 and accompanying text.

128 See, e.g., 5 U.S.C. § 5372a (contracts appeals board members); id. § 5372b (administrative appeals judges).

29

agency heads are sometimes governed by rules that require their impartiality in deciding cases, though the term is often undefined and may, of course, mean a number of different things. Some agencies—but by no means all—have recusal rules that, at their most robust, require adjudicators to recuse themselves from cases when their impartiality can reasonably be questioned. Recusal rules often do not cover agency-head adjudicators. But they may be governed by other recusal policies to root out conflicts of interest arising from prior private-sector affiliations.129

The research materials underlying a pending ACUS project provide valuable insights into the subjects addressed here and the difficulty of researching them.130 A committee-proposed recommendation asks agencies to disclose publicly on their websites specific information about adjudicators to help the public assess their impartiality and constitutional status.131

Finally, many adjudicators (other than ALJs) appointed by agency heads receive performance evaluations and can be compensated accordingly. They may be subject to case-volume requirements or goals. No program of which we are aware uses case outcomes as a performance metric, or at least none admit to doing so. Agencies generally disavow any reliance on case outcomes.

C. Selecting Cases for Appellate Review 1. Nature, Form, and Timing of Appeals. The general approach across

appellate programs is that a dissatisfied party—usually not the government, but sometimes that’s the case—can appeal a hearing-level agency decision. Similar to the federal court system, the first action is to file a notice of appeal, with the deadline set in the agency’s rules to do so generally ranging from a month to three months from the issuance of the hearing-level decision. One notable exception is the PTAB, where one must seek review within nine months of the patent’s issuance.132

129 See Admin. Conf. of the U.S., Recommendation 2018-4, Recusal Rules for Administrative

Adjudicators, 84 Fed. Reg. 2139 (Feb. 6, 2019); Louis J. Virelli, III, Administrative Recusal Rules: A Taxonomy and Study of Existing Recusal Standards for Agency Adjudicators (2020) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/administrative-recusal-rules-taxonomy-and-study-existing-recusal-standards-agency-0.

130 See Leigh Anne Schriever, Public Availability of Information About Agency Adjudicators (Oct. 21, 2020) (draft report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/draft-report-public-availability-information-about-agency-adjudicators. Attachments to the report include Kent Barnett, Some Kind of Hearing Officer, 94 WASH. L. REV. 515 (2019).

131 See Admin. Conf. of the U.S., Proposed Recommendation, Public Availability of Information About Agency Adjudicators (Nov. 6, 2020), https://www.acus.gov/recommendation/ recommendation-committee-adjudication-public-availability-information-about-agency.

132 App. L (PTO).

30

2. Standards for Granting (Discretionary) Review. Although some appellate programs review on the merits every appeal filed, others have discretionary appellate review. At the MSPB and SSA, for instance, the agencies have identified categories of cases or errors that merit administrative appellate review.133 In many ways, these criteria reflect the agency’s objectives in engaging in appellate review in the first place.

The SSA’s criteria, codified in regulation, is illustrative. The Appeals Council will review a case for five distinct reasons:

(1) There appears to be an abuse of discretion by the administrative law judge;

(2) There is an error of law; (3) The action, findings or conclusions of the administrative law

judge are not supported by substantial evidence; (4) There is a broad policy or procedural issue that may affect the

general public interest; or (5) Subject to paragraph (b) of this section, the Appeals Council

receives additional evidence that is new, material, and relates to the period on or before the date of the hearing decision, and there is a reasonable probability that the additional evidence would change the outcome of the decision.134

These criteria seem quite similar to setting the standard and scope of review, discussed below. They also may reflect the objectives the agency has set for its appellate review program.

Other agencies have a discretionary review process more similar to the U.S. Supreme Court certiorari review. At the SEC, for instance, the agency has clarified and simplified its certiorari process by requiring only a three-page petition to assist the Commission in deciding whether to hear the appeal on the merits.135 The PTAB, by statute, has a very formalized case institution process, where the PTAB institution panel assesses whether it is reasonably likely that the petitioner would prevail on at least one claim.136 The PTAB’s institution process has been subject to scrutiny at the Supreme Court. In SAS Institute, Inc. v. Iancu, the Court rejected the agency’s approach that it only reviewed on the merits the claims it certified at the institution stage.137 Instead, the Court held

133 See, e.g., App. J (MSPB); App. N (SSA). 134 20 C.F.R. § 404.970(a). 135 App. M (SEC). 136 App. L (PTO). 137 138 S. Ct. 1348, 1352 (2018).

31

that, once the PTAB decides to institute the case, the statute requires the agency to review every patent claim the petitioner has challenged.138

3. Agency-Initiated, Sua Sponte, and Interlocutory Review. The general approach in the appellate programs is to only review cases where a dissatisfied party has appealed. To be sure, some agencies have by statute or regulation the authority to review any hearing-level agency adjudication.139 For instance, the SEC has the authority to order review of any ALJ decision within 21 days of its issuance; only one commissioner is required to request such review.140 Yet as far as we are aware, this sua sponte discretionary review is seldom used at the vast majority of agencies, the unique non-adversarial SSA system (noted below) aside.

One notable exception is the Attorney General’s referral authority in the immigration adjudication context. As one of us has explored elsewhere, the Attorney General has the authority to sua sponte refer to himself any immigration adjudication case pending before the BIA and decide it himself.141 During the Trump Administration, the Attorney General has exercised this referral authority about a dozen times to set immigration policy via adjudication.142 Some immigration scholars and advocates have been critical of the Attorney General’s use of referral authority.143

Some agency appellate programs have instituted sua sponte review for quality assurance purposes. The most prominent example is the SSA.144 The Appeals Council uses random and selective sampling to identify cases for potential quality assurance review. A recent ACUS report on the use of artificial intelligence in the federal regulatory state examined the SSA’s use of these quality assurance review mechanisms.145 Similarly, at the SEC, the Office of

138 In dissent, Justice Ginsburg suggested that the PTAB could get around the Court’s holding by denying the petition on the claims that had no “reasonable likelihood” of success and then allowing the petitioner to amend or refile the petition with only the sufficient claims. Id. at 1360 (Ginsburg, J., dissenting). See generally Saurabh Vishnubhakat, Renewed Efficiency in Administrative Patent Revocation, 104 IOWA L. REV. 2643 (2019) (defending and further developing Justice Ginsburg’s approach in SAS Institute).

139 See, e.g., App. J (MSPB); App. K (NLRB); App. D (DHS). 140 App. M (SEC). 141 Wadhia & Walker, supra note 60, at Part I.A. 142 Id. 143 See, e.g., Bijal Shah, The Attorney General’s Disruptive Immigration Power, 102 IOWA L.

REV. ONLINE 129 (2017). Cf. Christopher J. Walker, Referral, Remand, and Dialogue in Administrative Law, 101 IOWA L. REV. ONLINE 84 (2016) (noting some advantages of the Attorney General’s referral authority).

144 App. N (SSA). 145 David Freeman Engstrom, Daniel E. Ho, Catherine M. Sharkey & Mariano-Florentino

Cuéllar, Government by Algorithm: Artificial Intelligence in Federal Administrative Agencies 37–45 (2020) (report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/government-algorithm-artificial-intelligence-federal-administrative-agencies; see also Gerald K. Ray &

32

Adjudications within the Office of General Counsel randomly reviews decisions not appealed.146 The SSA Appeals Counsel also reviews cases on its own motion or based on referral. Other appellate programs likewise allow a hearing-level adjudicator to request review of a pending case.147

D. Decision-Making Process and Decisions 1. Multi-Adjudicator Decision Making. Similar to the federal court

system, most agency appellate programs use multi-member bodies to make decisions. The NLRB, MSPB, and SEC, for example, hear many appeals as a full body—though the MSPB’s process is for the case to start with one board member and then circulate to each other member one at a time.148

With respect to larger appellate bodies, a select number of appellate adjudicators sit on each panel. At the PTAB, three administrative patent judges generally sit on each merits panel.149 At the BIA, merits panels are similarly composed of three members, but the BIA also handles many appeal denials through one-member decisions.150

On the SSA Appeals Council, only one appellate adjudicator is required to deny an appeal, but two are required to grant.151 With the assistance and recommendation of a staff attorney, if one adjudicator decides to grant, the case record and draft decision are transmitted to a second adjudicator. If she agrees, that becomes the final decision. If she disagrees, the case is sent to a third adjudicator. The dissenting adjudicator’s vote (as well as the staff attorney’s recommendation) are not included in the administrative record or noted in the final decision.

Not all agency appellate review programs, however, use multi-member panels. The USCIS and HHS, for example, use the one-judge model, with the decisions from the USCIS all signed under the name of the head of the appellate body.152

During the interviews, the agency officials expressed many views on the costs and benefits of these different approaches to staffing appellate programs. Most seemed to agree that the multi-member approach is the most beneficial for complex cases. Some, however, recommended the single-member approach for simple cases and screening purposes. And others indicated that resource Jeffrey S. Lubbers, A Government Success Story: How Data Analysis by the Social Security Appeals Council (with a Push from the Administrative Conference of the United States) is Transforming Social Security Disability Adjudication, 83 GEO. WASH. L. REV. 1575 (2015).

146 App. M (SEC). 147 See, e.g., App. C (HHS); App. D (DHS). 148 See, e.g., App. K (NLRB); App. J (MSPB); App. M (SEC). 149 App. L (PTO). 150 App. E (DOJ). See generally Hausman, supra note 53, at 1205–07. 151 App. N (SSA). 152 App. C (HHS); App. D (DHS).

33

constraints limited their ability to use multi-member review panels when they would otherwise be ideal. Finally, even when an agency only uses the one-member review approach, the officials underscored that a number of agency officials are often involved in the process. The use of staff attorneys and law clerks is discussed below.

2. Docket/Case Management. During the interviews, agency officials underscored the various innovations developed in their appellate programs to improve docket and case management. Many agencies have implemented mechanisms to screen and sort cases based on a variety of factors, including the degree of difficulty, the subject matter, common issues, record length, geographical area (especially where there are significant variations in circuit law), the need for additional evidence or oral argument, or even likely outcome.153 Some agencies—like the SSA and HHS—have experimented with using automation, advanced data analytics, and artificial intelligence to help sort cases. Many rely on staff attorneys to help with docket management as well as case screening and sorting.

Similarly, a number of agencies organize their appellate adjudicators based on adjudicator expertise, familiarity with a common issue, or difficulty of case. Some agencies expedite certain appeals based on various factors as well as group like cases to be decided at the same time—instead of following a more conventional first-in, first-out docket management approach. ACUS has previously published an extensive study on aggregate agency adjudication, which provides in-depth case studies on a number of such innovations.154

3. Use of Staff Attorneys and Law Clerks. During the interviews, many agency officials offered as a best practice their agency’s use of staff attorneys, paralegals, law clerks, and other expert staff to help with docket management and screening as well as with case review, decision recommendation, and opinion drafting. At many agencies, these staff attorneys and law clerks are integrated into the appellate bodies—either working for the appellate body generally or for a specific adjudicator in particular.155 Some staff attorneys and clerks specialize in certain issues, whereas others are randomly assigned to work on matters. At the MSPB and SEC, by contrast, the agency has a separate office—the Office of Appeals Counsel and the Office of Adjudications, respectively—where the staff attorneys work to assist the adjudicators.156

153 See, e.g., App. C (HHS); App. D (DHS). 154 Michael Sant’Ambrogio & Adam Zimmerman, Aggregate Agency Adjudication (2016)

(report to the Admin. Conf. of the U.S.), https://www.acus.gov/report/aggregate-agency-adjudication-final-report.

155 See, e.g., App. G (VA); App. H (EPA); App. K (NLRB); App. L (PTO). 156 App. J (MSPB); App. M (SEC).

34

4. Standard of Review and Submission of New Evidence. Many agency appellate programs have adopted the federal court’s appellate review model, discussed in Part I.C, with respect to the scope and standard of review.157 In other words, the appellate body reviews a closed administrative record, and the standard of review varies based on the relevant competence of the trial- and appellate-level adjudicator. Some appellate programs conduct de novo review, often with a narrow exception for deference to hearing-level credibility determinations.158 Other agencies review factual findings under a substantial-evidence standard.159 The APA and most other statutes leave the choice to agencies.160

One way in which agency appellate bodies (even those that have embraced the appellate model) differ from the federal courts of appeals concerns the submission of new evidence. A number of agencies that otherwise function as traditional appellate bodies in most respects allow the submission of new evidence on appeal when there is good cause rather than remanding the case to the hearing-level adjudicator to receive the evidence.161

But, at least so far as the submission of new evidence is concerned, some agencies have not adopted the appellate review model at all.162 Instead, they allow for the introduction of new evidence on appeal and, in essence, conduct a trial de novo. The USCIS is a classic example, where the current regulations allow for a trial de novo on appeal. In December 2019, however, the USCIS filed a notice and sought public comment on changes to its scope and standard of review.163 In that notice, the agency proposed to only allow new evidence on appeal that was not reasonably available in the hearing-level adjudication and to review discretionary decisions for abuse of discretion. Another example (allowing for some qualifications) is the BVA.164

During the interviews, many agency officials emphasized the importance of setting the right scope and standard of review for their appellate programs. The appropriate scope and standard, some noted, depends on the objectives the agency seeks to maximize via appellate review of hearing-level adjudications.

157 See, e.g., App. E (DOJ); App. L (PTO). 158 See, e.g., App. J (MSPB); App. K (NLRB); App. H (EPA). 159 See, e.g., App. N (SSA); App. I (EEOC). 160 See, e.g., 5 U.S.C. § 557. 161 See, e.g., App. I (EEOC); App. J (MSPB); App. K (NLRB); App. M (SEC); App. N (SSA). 162 See, e.g., App. G (VA); App. C (HHS) (some cases); App. D (DHS). 163 U.S. Citizenship & Immigration Servs., Agency Information Collection Activities;

Revision of a Currently Approved Collection: Notice of Appeal or Motion, 84 Fed. Reg. 66924 (Dec. 6, 2019), https://www.govinfo.gov/content/pkg/FR-2019-12-06/pdf/2019-26331.pdf.

164 See App. G (VA); see also ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 180. The VA has sought legislation to limit the introduction of new evidence on appeal. See id.

35

5. Oral Argument. In accord with the appellate model in federal courts, most agency appellate programs allow for the possibility of oral argument, usually at the request of a party or at the instigation of the appellate body itself.165 Most agency appellate programs, like the federal courts of appeals, retain discretion whether to grant oral argument. Few agencies’ procedural rules, or even guidance documents, set forth any specific criteria for exercising that discretion.166 Some agencies grant it more frequently than others. Some seldom use it. Many use videoconference technology as one means for conducting oral argument.

6. Amicus Participation. Many agency appellate programs allow amicus participation, but that is not always clear from their practice manuals or rules. Some appellate bodies, like the NLRB, sometimes solicit amicus briefs in significant cases. At the MSPB, for instance, the Office of Personnel Management and the Office of Special Counsel can be asked to weigh in, and in some cases have the statutory right to do so.167 In immigration adjudication, whether to allow amicus briefs is at the sole discretion of the BIA.168 At the USCIS, amicus briefs are only allowed if they are solicited by the party or the USCIS.169 Amicus briefs can be especially important at agencies like the NLRB that rely heavily on adjudication for policymaking, which may require consideration of legislative facts not present in the hearing-level record.

7. Deadlines for Decisions. Most agencies do not impose deadlines for appellate decisions—by statute, regulation, or rules of practice. But there are a few exceptions. For example, the HHS has a 90-day deadline for Medicare appeals.170 The PTAB has a one-year deadline for institution, which can be extended up to six months for good cause.171 At the SEC, there is no statutorily imposed deadline, but the Rules of Practice encourage appellate decisions within eight months.172 And at the USCIS, the appellate body “strives” to complete its review within 180 days of receiving the complete case record.173

During the interviews and as discussed further below, many agency officials observed that backlogs and delays were a significant challenge in their appellate review programs. Yet, at the same time, those officials were not confident that

165 Cf. Envtl. Prot. Agency, Envtl. Appeals Bd., 40 C.F.R. § 124.19(h) (providing for presumption against oral argument).

166 But see, e.g., Dept. of Labor, Medicare Appeals Council, 42 C.F.R. § 1124(a)) (providing general criteria).

167 App. J (MSPB). 168 App. E (DOJ). 169 App. D (DHS). 170 App. C (HHS). Ongoing noncompliance with this deadline in the face of intractable

backlogs has resulted in extensive litigation. See, e.g., Am. Hosp. Ass’n v. Price, 867 F.3d 160 (D.C. Cir. 2017); Cumberland Cty. Hosp. Sys., Inc. v. Burwell, 816 F.3d 48 (4th Cir. 2016).

171 App. L (PTO). 172 App. M (SEC). 173 App. D (DHS).

36

a mandatory decision deadline would help address those problems. Instead, the officials emphasized the need for more agency adjudicators and staff and for technological advances, such as electronic case management and innovations in the use of automation and artificial intelligence.

8. Forms of Decisions. In many ways, the forms of appellate decisions at federal agencies are similar to the forms of appellate decisions in federal courts. Agency appellate bodies can issue decisions that affirm, reverse (and grant relief or remand), or vacate and remand the case to the hearing-level adjudicator. This can involve detailed published or unpublished decisions, or shorter orders that summarily affirm the hearing-level adjudicator’s decision.

Yet in reviewing appellate decisions from the various agencies, it immediately becomes apparent that there is great diversity in terms of opinion structure, format, style, readability, and overall quality. Some agency appellate decisions, like those at the SEC, read very much like a published, precedential opinion from a federal court of appeals. Others, like at the NLRB, often adopt and incorporate at length the underlying ALJ decision and then just note and discuss the exceptions to the ALJ decision. And others, to put it charitably, are very difficult for a non-lawyer to understand. This is one area where some agencies could strive to make their appellate programs more accessible to the appellants in the system—some of whom have no legal representation. Clarity and brevity, of course, should be the aspiration of any decision writer. One agency’s appellate rules of practice actually mandate brevity in opinions.174

During the interviews, many agency officials emphasized measures their agency has taken to write opinions in plain language and to use templates and standard language to improve the quality of decisions. For example, the BVA has implemented an Interactive Decision Template (IDT), which “automatically retrieves data from the Board’s case management software and allows attorneys to populate important and relevant language into each appellate decision, allowing attorneys and VLJs to focus their attention on legal research and drafting. The IDT also ensures consistency and brevity of all Board decision and encourages the use of clear and concise language to ensure all Board decisions are understandable to Veterans.”175 As some agencies have found on judicial review, agency templates must be carefully drafted to accurately reflect governing law and account for case-specific issues that may require variances.176

9. Precedential Designation. Some agencies do not distinguish between precedential (sometimes called published) and nonprecedential (sometimes called unpublished) decisions. That is most often the case with the decisions of

174 See Envtl. Appeals Bd., 40 C.F.R. § 124,19(l)(3). 175 U.S. DEP’T OF VETERANS AFFAIRS, BOARD OF VETERANS’ APPEALS, ANNUAL REPORT 17 (FY

2018), https://www.bva.va.gov/docs/Chairmans_Annual_Rpts/BVA2018AR.pdf. 176 One agency’s templates, now since modified, came under especially harsh criticism in the

courts. See, e.g., Bjornson v. Astrue, 671 F.3d 640 (7th Cir. 2012) (per Posner, J.).

37

agency heads, especially boards and commissions like the NLRB. All of their decisions may have precedential status—whether by explicit rule,177 decisional law, long-accepted practice, or perhaps even the judicial requirement of reasoned decision-making across cases178—with only small exceptions sometimes made for summary affirmances and other short-form dispositions. Many agencies, though, have increasingly adopted the practice of designating only certain decisions—often few in number—as precedential.

There is great variation among agency appellate programs about whether and how to designate an appellate decision as precedential. For some agencies, their appellate programs are not designed to make policy, and thus the agencies, rightly or wrongly, perceive no need for precedential opinions. At the BVA, for instance, when the appellate body wants to make policy, it works with the VA’s Office of General Counsel to bring consistency to adjudications.179 The USCIS seldom issues precedential decisions because the process is lengthy and resource intensive, requiring signoffs from the DHS General Counsel, the Justice Department’s Office of Legal Counsel, and the Attorney General, and then publication by the BIA.180

Other agencies engage more often in publishing precedential decisions. At the SEC, opinions are generally precedential and published, as the Commission routinely makes policy through adjudication.181 In immigration adjudication, the vast majority of BIA decisions are nonprecedential and unpublished. But the BIA can and does go through an en banc process to select decisions to publish as precedential.182 As discussed above, immigration adjudication decisions by the Attorney General are also precedential. During the Trump Administration, the BIA and the Attorney General have issued around one-hundred precedential decisions.183

As one of us has detailed elsewhere, the PTAB had a very difficult process for designating decisions as precedential.184 In 2018, however, the PTAB adopted new procedures, which established a precedential opinion panel and set forth procedures to identify cases for this new panel to consider making precedential.185

For agencies that seek to make policy through adjudication, the use of precedential opinions has great promise. Indeed, even for those that do not seek

177 See, e.g., App. J (MSPB). 178 See, e.g., Allentown Mack Sales & Serv., Inc. v. NLRB, 522 U.S. 359, 374–77 (1998). 179 App. G (VA). 180 App. D (DHS). 181 App. M (SEC). 182 App. E (DOJ). 183 Wadhia & Walker, supra note 60, at Part II.B. 184 See Walker & Wasserman, supra note 26, at 188–96. 185 App. L (PTO).

38

to make policy via adjudication, the use of precedential opinions has the potential to bring greater consistency and uniformity to their administrative adjudication programs186—an often-overlooked value—as the PTAB and a few other agencies have explicitly noted in their rules setting for the criteria for designating decisions as precedential.187 But many agencies lack clear procedures and criteria for publishing appellate decisions as precedential, and others may lack the statutory or regulatory authority to even engage in this practice.

One important procedural question suggested above is where and when the authority to designate decisions as precedential lies. At some agencies, the adjudicators who decide particular cases have that authority, which they exercise incidental to deciding the case, much like a panel of a federal court of appeals. This practice is most common when the decision-maker is a board or commission.188 It aligns very much with the judicial model of appellate review.

At other agencies, by contrast, the decision whether to designate a decision as precedential is made after its issuance by an official other than the issuing adjudicator—sometimes through a bureaucratic-like process that sets it apart from the judicial model. In addition to the USCIS process described above, examples include the Board of Immigration Appeals, at which a panel decision is designated as precedential on the vote of the majority of the full Board (or at the direction of the Attorney General);189 the Medicare Appeals Council, at which a panel decision is designated as precedential by the chair;190 the Environmental Appeals Board, at which the Board’s designation of a decision as precedential (“published”) takes effect only after 15 days so as to give the EPA administrator the opportunity to reverse the precedential designation; and, most notably, the PTAB, at which a panel decision is designated as a precedential by a special panel selected and convened by the PTO’s director.191 The PTAB’s process is probably the most formalized and elaborate.

Several appellate programs have adopted processes under which they solicit suggestions—from other adjudicators, agency officials, litigants, and the

186 See ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 85. 187 See, e.g., Patent Tr. & Appeal Bd., Standard Operating Procedure No. 2 (rev. 10) (noting

that the “Director has an interest in creating binding norms for fair and efficient Board proceedings, and for establishing consistency across decision makers”), (https://www.uspto.gov/ sites/default/files/documents/SOP2%20R10%20FINAL.pdf (last visited Dec. 1 ,2020); Dept. of Justice (Bd. of Immigr. Appeals), 8 C.F.R. § 1003.1(g) (stating that purpose of precedential decisions is to bring “consistency and uniformity” to adjudication and enforcement decisions).

188 See Merit Sys. Protection Bd., 5 C.F.R. § 1201.117(c). 189 See 8 C.F.R. § 1003.1(g)(3). 190 App. C (HHS). 191 App. L (PTO).

39

public—for precedential designations.192 The PTAB does so through an online-form.193

10. Reconsideration. Virtually all agency appellate programs have some mechanism for reconsideration, re-opening, en banc review, or agency-head review of agency appellate decisions. The criteria for such further review vary by agency. Unsurprisingly, reconsideration is seldom granted, and is usually reserved for when there is new evidence, egregious error, or some other extraordinary circumstance.

11. Judicial Remand and Court-Agency Dialogue. Although not abundantly apparent in the publicly available information, our interviews with agency officials shed fascinating light on the interaction between agency appellate bodies and federal courts. It turns out that in most agency appellate programs, adjudicators pay close attention to judicial decisions that remand cases to the agency. And they take a number of measures to interact with the judicial decision on remand.

Most obviously, the agency on remand must issue a new decision that responds to the federal court’s decision. (Some agencies also have the ability to settle cases on remand.) To do this, many appellate bodies confer with their full membership to discuss how to respond. Among potential responses, the appellate adjudicators must often decide when to adopt the federal court’s precedent nationwide; merely acquiesce to the precedent in the relevant federal circuit court’s jurisdiction (or not); or, in the statutory interpretation context, perhaps even adopt a different interpretation and seek Chevron deference under the Brand X doctrine.194 Sometimes the appellate adjudicators will decide to remand the case to the hearing-level adjudicator to make new factual findings or otherwise reconsider the remanded issue in the first instance.

It is important to note that the U.S. Court of Appeals for the Federal Circuit has exclusive jurisdiction for three of the appellate programs in our study: BVA, MSPB, and PTAB. (Only a small class of MSPB cases may be appealed to a circuit other than the Federal Circuit.) For those agencies, there is no decision whether to acquiesce. But officials from those agencies underscored how deep and interactive their relationship with the Federal Circuit has become. The agency appellate body and the Federal Circuit are in a continuing dialogue about the development of policy and precedent and the functioning of the adjudication program. Officials at agency appellate bodies that are regularly

192 See, e.g., id.; App. C (HHS) (Medicare Appeals Council). 193 Precedential Opinion Panel (POP) Amicus Form, https://www.uspto.gov/patents-

application-process/patent-trial-and-appeal-board/precedential-opinion-panel-pop-amicus (last visited Dec. 1, 2020).

194 See Nat’l Cable & Telecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 982 (2005) (“A court’s prior judicial construction of a statute trumps an agency construction otherwise entitled to Chevron deference only if the prior court decision holds that its construction follows from the unambiguous terms of the statute and thus leaves no room for agency discretion.”).

40

reviewed by the U.S. Court of Appeals for the D.C. Circuit made similar observations about the special relationship they have with the D.C. Circuit.

Federal agencies do not only respond to judicial remand through subsequent agency decisions. They also respond by issuing guidance and training within the agency—for the hearing-level and appellate adjudicators and for other agency officials. At some agencies, the appellate body takes the lead in this training. At others, the agency general counsel’s office plays that role. A number of officials interviewed noted how the agency appellate bodies discuss these judicial decisions at regular meetings and agency-wide conferences and trainings. In other words, the judicial remand decisions have the potential to have a much more systemic effect on agency operations and adjudication outcomes.

One final note: The agency officials interviewed underscored that this court-agency interaction is not a one-way street. It is not just the circuit court that influences the agency; the agency’s decisions also influence the court’s approach. They view this interaction as more of a partnership than a supervisory relationship—and as more of a dialogue than a judicial monologue. We return to this subject in Part III.G. E. Administration, Management, and Bureaucratic Oversight

The managerial and bureaucratic side of agency appellate programs, like that of any adjudication program, is immensely important and could consume multiple volumes. Below we address several aspects of the subject that lend themselves to cross-agency consideration, though not necessarily the formulation of best practices.

1. Guidance Documents Directed to Adjudicators. Adherents to the judicial model, as opposed to a bureaucratic/managerial-type model, of agency adjudication tend to see the legal authorities governing hearing-level and appellate decision making as mostly limited to statutes; legislative rules, whether substantive and procedural; and decisional law, both agency and judicial. (Other legal pronouncements at certain agencies might be added to this list.195) They tend to reject internal administrative law—in particular, interpretive rules and policy statements, which together are now often put under the rubric of “guidance documents”196—as being authoritative.

Our assessment is borne out by the agency appellate programs under review. Few use guidance documents to direct the actual decision making of hearing-level adjudicators, or appellate adjudicators for that matter. (Internal policies, of course, govern all sorts of administrative and other matters that are critically

195 See, e.g., 38 U.S.C. 7104 (providing that the BVA’s decisions are governed not only by

regulations, but also “instructions of the Secretary, and the precedent opinions of the chief legal officer of the Department”).

196 See, e.g., Levin, supra note 102.

41

important to appellate programs.) The arguable exception here are interpretive rules issued by agency heads.197

The obvious outlier here, not surprisingly, is the SSA Appeals Council.198 The Appeals Council’s appellate disposition of ALJ decisions consists of brief, case-specific, unpublished (on SSA’s website or otherwise), and nonprecedential decisions. They do not establish a body of decisional law (substantive or procedural) that governs the decisions of SSA ALJs generally. When the Appeals Council speaks to ALJs outside the context of particular cases, it does so mostly through the previously mentioned HALLEX manual199—a guidance document that exemplifies internal administrative law.200 HALLEX addresses mostly procedural matters that arise in ALJ hearings as well as proceedings before the Appeals Council. One litigated provision, for example, tells an ALJ when a translator must be provided rather than leaving it up the ALJ’s case-by-case determination.201

The SSA disclaims any status of HALLEX as a legislative rule—that is, a rule that a litigant could enforce on judicial review of an SSA decision202—and to this end does not publish it in the Federal Register or the CFR.203 But the SSA deems HALLEX binding on ALJs—on pain of discipline through the MSPB204—and will use its provisions to assign error to an ALJ decision on appeal.

An especially notable aspect of HALLEX as far as this study is concerned is what might be called its interpretive components. They take the form of Appeals Council Interpretations: short interpretations of SSA law on discrete issues which purposes include “resolv[ing] issues arising from gaps in policy or unclear statements of policy to facilitate the adjudication of individual cases coming before the Council for consideration.”205 The impetus for an interpretation may

197 App. M (SEC). 198 See generally Nicholas R. Parrillo, Introduction, in ADMINISTRATIVE LAW FROM THE

INSIDE OUT, supra note 10, at 6 (questioning whether the extensive use of “internal law” in SSA’s adjudication system, as advocated by Professor Mashaw in his pathbreaking scholarship, is “replicable beyond SSA”).

199 The SSA Commissioner, as distinct from the Appeals Counsel, also issues Social Security Rulings (SSRs) and Acquiescence Rulings, both guidance documents. The latter, as the name suggests, sets forth the SSA’s position as to when the agency will acquiesce in circuit-court law at variance with SSA policy in cases arising within the particular circuit. See Soc. Sec. Admin., Social Security Rulings and Acquiescence Rulings, https://www.ssa.gov/OP_Home/rulings/ rulings-toc.html (last visited Nov. 26, 2020).

200 See Part III.A.2 supra. 201 See HALLEX, supra note 108, § I-2-6-10. 202 See Part III.A.2 supra. 203 Nor has the SSA ever noticed HALLEX for public comment. Much of it, being procedural,

would be exempt from the APA’s notice-and-comment requirements, although the SSA generally does not avail itself of that exemption even when it promulgates entirely procedural rules.

204 See, e.g., Social Sec. Admin. v. Butler, 2016 WL 4547551 (M.S.P.B. 2016). 205 HALLEX, supra note 108, § II-5-0-1.

42

be a particular case, but the interpretation is, and is styled as, a guidance document rather than a case-specific decision. HALLEX establishes formal procedures for the Appeals Council’s issuance of interpretations.206

One reason agencies—and, in particular, non-agency-head appellate adjudicators—may shy away from using guidance documents to direct the decision making of hearing-level adjudicators is that the practice can raise questions as to whether, when, and how they bind adjudicators. In its recent recommendation on the internal bindingness of guidance documents, ACUS itself avoided the issue as to whether agency adjudicators are among the agency employees to whom “direction” can be given by policy statement.207 Our own, and we suspect the predominant, view is that agencies may do so in many instances.

It may be that appellate programs do not really need to rely on guidance documents to direct the decisional activities of their adjudicators and promote inter-decisional uniformity and consistency. Precedential decisions, when well-drafted, may serve the same objectives, and do so in a way more consonant with the judicial model of agency adjudication and the expectations of at least some agency adjudicators. That is especially true when appellate programs effectively communicate precedential decisions to adjudicators and facilitate their access to them through summaries, digests, and other mechanisms discussed below.

2. Feedback to Hearing-Level Adjudicators. The SSA Appeals Council aside, none of the appellate programs we studied provide direct feedback in any form to hearing-level adjudicators on their decisions overall. The feedback comes in the form of an appellate decision reviewing a hearing-level adjudicator’s decision in an individual case. Occasionally an appellate decision notes a recurring practice of the particular hearing-level adjudicator that needs correction. Having noted such a problem in a decision, the appellate decisionmaker usually leaves the matter in the hands of the hearing-level office to address with the adjudicator, if it is to be addressed at all. Here again the judicial model of agency adjudication dictates agency practice.

3. Dialogue with Rulewriters. A recurring issue in administrative law is the choice between rulemaking or adjudication to make policy.208 Appellate adjudicators are often uniquely positioned to identify recurring issues in adjudications, both procedural and substantive, that might be best addressed by rules. When the appellate adjudicator(s) is the agency head, as in the case of a multi-member board or commission, the adjudicators are obviously well positioned to consider, and then initiate, possible rulemaking initiatives.

206 See id. 207 See Admin. Conf. of the U.S., Recommendation 2019-1, Agency Guidance Through

Interpretive Rules, 84 Fed. Reg. 38,927 (Aug. 8, 2019); Admin. Conf. of the U.S., Recommendation 2017-5, Agency Guidance Through Policy Statements, 82 Fed. Reg. 61,734 (Dec. 29, 2017).

208 The literature is voluminous. See, e.g., Magill, supra note 60; Shapiro, supra note 60.

43

Witness recent rulemaking initiatives (albeit substantively controversial ones) at the NLRB.209

We found that, when an appellate program resides below the level of the agency head, dialogue between adjudicators and rulewriters is often sporadic and ad hoc, when it occurs at all. One reason at some agencies appears to be that adjudicators, socialized into the judicial review model of agency adjudication discussed in Part I.C, do not consider policymaking to be their concern. Some even think it to be incompatible with their adjudicative roles. Another reason is administrative. Agency organizational structures sometimes inhibit dialogue and collaboration between adjudicators and rulewriters.

We do not mean to imply that this problem is characteristic of all appellate programs. Some agencies have done admirably well in facilitating regular dialogue between rulewriters and adjudicators. The SSA is the most notable example. It has even adopted a formal, though infrequently used, mechanism—the above-mentioned Appeals Council Interpretations—to “identify[] and surfac[e] to SSA policymakers conflicts or inconsistencies in program policy.”210 But agencies like the SSA appear to be the exception.

As we note in the Conclusion, ACUS may wish to further study this issue and, insofar as administrative best practices can be identified to institutional adjudicator-rulewriter dialogue, embody them in a recommendation.

4. Notifying Adjudicators of Agency Appellate Decisions and Judicial Decisions. Some appellate programs actively notify both appellate and hearing-level adjudicators of their decisions and judicial decisions rather than relying on the adjudicators to independently review them from the program’s website or otherwise. What decisions a program notifies agency adjudicators of and how it does so vary considerably.

As for their own decisions, agency appellate programs most often see the need to notify appellate adjudicators when the program has a high-volume caseload with many adjudicators, only one or a small number of whom (usually sitting in a panel) actually participate in any given case. Notification is especially important—and, in a high-volume program, often appropriately reserved for—precedential and other important decisions of which adjudicators must be aware when deciding future cases. (Notification would be impracticable and often not very useful in high-volume programs where decisions involve routine application of well-established law and policy—procedural and substantive—to case-specific facts.) No generalization can be made about the optimal format and means of delivery. Some agencies find that sending

209 See Nat’l Labor Relations Bd., National Labor Relations Board Rulemaking,

https://www.nlrb.gov/about-nlrb/what-we-do/national-labor-relations-board-rulemaking (last visited Nov. 26, 2020).

210 HALLEX, supra note 108, § II-5-0-1.

44

memoranda suffices. Agencies often find it useful, not surprisingly, to provide summaries and commentary about the decisions.

As for judicial decisions, most agency appellate programs notify their adjudicators about them or, in high-volume programs that result in a high-volume of judicial opinions, at least some of them. High-volume programs are more likely to notify adjudicators about circuit-court rather than district-court opinions. Here, again, such notifications are often best accompanied by summaries and brief commentary.

Judicial decisions do differ from agency appellate decisions in one important respect that bears on the issue of notice: Supreme Court decisions aside, federal-court decisions are not usually binding on agencies except as to the law of the case on review. The issue is especially pronounced when circuits split on a point of law. A few agencies have found it necessary to provide guidance to their adjudicators as to when they should follow federal-court decisions and when agency policy remains unaffected.

The foregoing discussion has addressed only an appellate program’s notification to its own appellate adjudicators. Direct notification of hearing-level adjudicators—and, as we note below, direct communication with them generally—appears uncommon. Appellate programs usually leave notification to hearing-level administrators. This is one context in which any number of agencies could improve dialogue, and at least some modest coordination, between appellate-level offices and hearing-level offices.

5. Digests and Indexes of Decisions. Several of the studied agencies maintain digests and indexes of agency and sometimes judicial decisions for hearing-level and appellate adjudicators. This facilitates necessary research and is especially important in adjudication programs in which decisional law (as opposed to statutory or rule-based law) plays an important role in deciding even routine cases. Some agencies post short-form versions of digests and indexes on their website so that both adjudicators and the public can benefit from them.211

Digests of agency precedents are especially useful when adjudication decision making requires substantial engagement with an extensive body of doctrinally complex agency precedent. Some agencies could make more extensive use of them.

A still more ambitious mechanism that agencies might consider employing—though none, as far as we know, have employed so far—would be internally prepared restatements of the substantive law governing decisions with extensive notes or comments and citations. (The closest analogue are “bench books,” but they mostly concern procedural matters and emanate from the hearing level.) Such restatements might save appellate and hearing-level adjudicators time and prevent avoidable doctrinal errors. Most agencies, we

211 See examples cited in notes 221–226 infra.

45

suspect, would probably want to disclaim any intent to bind adjudicators or establish judicially enforceable law through such restatements.

6. Electronic Case Management and Filing. Many but by no means all appellate programs use an electronic case management system (eCMS). Some programs’ transition from paper recordkeeping systems is only partially complete, and some still rely substantially on paper records.

ACUS has already addressed that subject at length for adjudication programs generally.212 We have only one additional observation drawn from the practices of several of the agencies we studied: an eCMS for appellate programs should, as far as practicable, be integrated with the eCMS for hearing-level programs to facilitate an appellate adjudicator’s access to hearing-level filings, decisions, and (especially) the evidentiary record.

A related issue is electronic filing. Most appellate programs, like hearing-level programs, provide for electronic filing.213 Programs that have not yet done so should do so as promptly as resources allow.

7. Video Hearings. ACUS has already addressed the use of video hearings for adjudications214 and recently began a successor project on a related subject.215 Moreover, ACUS has recommended that hearing-level proceedings be presumptively open to the public, potentially with streamed or recorded audio or video.216 Our research and interviews suggest that these recommendations should generally govern appellate proceedings, especially oral argument.

8. Interagency Dialogue. One overarching conclusion we can confidently draw from our interviews is that most agencies seldom exchange information and perspectives on the matters in this subsection with other agencies. This inter-agency silo effect is an impediment to optimal administration. ACUS is uniquely situated to address this problem—not only through the formal recommendation of its Assembly but also through the information-exchange functions undertaken by its Office of the Chairman under the Administrative

212 See Admin. Conf. of the U.S., Recommendation 2018-3, Electronic Case Management in

Federal Administrative Adjudication, 83 Fed. Reg. 30,686 (June 29, 2018); see also ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 74.

213 See ASIMOW, ADJUDICATION SOURCEBOOK, supra note 20, at 78–80. 214 See Admin. Conf. of the U.S., Recommendation 2014-7, Best Practices for Using Video

Teleconferencing for Hearings, 79 Fed. Reg. 75,114 (Dec. 17, 2014); Admin. Conf. of the U.S., Recommendation 2011-4, Agency Use of Video Hearings: Best Practices And Possibilities For Expansion, 76 Fed. Reg. 48,795 (Aug. 9, 2011).

215 See Admin. Conf. of the U.S., Virtual Hearings in Agency Adjudication, https://www.acus.gov/research-projects/virtual-hearings-agency-adjudication (last visited Nov. 26, 2020).

216 Recommendation 2016-4, supra note 30.

46

Conference Act.217 One modest, but important, step the Office has recently taken to enhance this function is the establishment of its Council on Federal Agency Adjudication.218 This Council should address issues in appellate programs in addition to hear-level adjudication programs.

9. Solicitation of Public Feedback. Several of the studied agencies solicit public feedback on their procedures and processes. In Recommendation 2016-4, ACUS recommended this practice for hearing-level adjudication programs.219 This recommendation applies with similar force to appellate programs.

F. Public Disclosure and Transparency 1. Adjudication Materials on Agency Websites Generally. ACUS has

adopted two recommendations on adjudication materials (rules of practice, decisions, explanatory materials, and so forth) on agency websites.220 Here are a few additional observations on the subject particular to appellate programs drawn from our review of the studied agencies’ appellate programs and other websites:

a. Subject to resource constraints, appellate programs should consider putting all their appellate decisions on their websites, unless they are so voluminous and case-specific as to be of little public interest (as might be the case with SSA Appeals Council decisions). This includes even high-volume adjudication programs. The BVA, most notably, has shown that this can be done.221 Of course, it may be appropriate or legally necessary to redact personally identifiable, confidential, private, and other such information. A number of agency websites offer examples.222

b. The best websites allow the public to search decisions for key words and allow the public to sort agency decisions by name, date, docket number, and

217 See 5 U.S.C. §§ 591, 594(2), 595(b); cf. Jennifer Nou, Agency Coordinators Outside of the Executive Branch, 128 HARV. L. REV. FORUM 64 (2015) (discussing possible role for ACUS in inter-agency coordination involving adjudication).

218 Press Release, Admin. Conf. of the U.S., ACUS Announces New Council on Agency Adjudication (Sept. 10, 2020), https://www.acus.gov/newsroom/news/acus-announces-new-council-federal-agency-adjudication.

219 See Recommendation 2016-4, supra note 30, § 31. 220 See Recommendation 2018-5, supra note 104; Admin. Conf. of the U.S., Recommendation

2017-1, Adjudication Materials on Agency Websites, 82 Fed. Reg. 31,039 (July 5, 2017). The report associated with the latter is Daniel J. Sheffner, Access to Adjudication Materials on Federal Agency Websites, 51 AKRON L. REV. 447 (2017).

221 See U.S. Dep’t of Veterans Affairs, The Board of Veterans’ Appeals Decision Search Results, https://www.index.va.gov/search/va/bva.jsp (last visited Nov. 26, 2020).

222 See, e.g., Equal Emp’t Opportunity Comm’n, The DIGEST of Equal Employment Opportunity Law, https://www.eeoc.gov/federal-sector/digest/digest-equal-employment-opportunity-law (last visited Nov. 26, 2020); U.S. Citizenship & Immigration Servs., AAO Non-Precedent Decisions, https://www.uscis.gov/administrative-appeals/aao-decisions/aao-non-precedent-decisions (Nov. 26, 2020).

47

other criteria. Allowing sorting by subject matter is an especially useful feature.223 The NLRB’s website suggests another best practice that may be appropriate for some agencies: listing especially significant decisions with accompanying brief summaries prepared by agency staff.224 One appellate program, the BIA, goes further and includes with each precedential decision in its bound volumes (now online) a one-sentence summary of the decision’s principal holdings (somewhat akin to the syllabus that accompanies Supreme Court decisions).225

c. Most agencies that distinguish between precedential and nonprecedential decisions follow the practice of listing them separately under clear headings.226 An accompanying explanation of the difference, conspicuously displayed, is also helpful.227 Not all agencies follow that practice.

d. Some agencies put on their websites notices of recent decisions, often accompanied by useful, short summaries. One agency’s website we reviewed helpfully organizes recent decisions by subject matter.228 It is probably good practice to accompany summaries by a disclaimer that the agency is providing them only for informational purposes and that they do not have any legal effect, although it is hard to imagine a court concluding otherwise.229

223 Envtl. Prot. Agency (Env’t Appeals Bd.), Published Decisions [by] Lists of Statutes,

https://yosemite.epa.gov/oa/EAB_Web_Docket.nsf/Statutes?OpenPage (last visited Nov. 26, 2020).

224 See Nat’l Labor Relations Bd., Cases and Decisions—Notable Board Decisions, https://www.nlrb.gov/cases-decisions/decisions/notable-board-decisions (last visited Nov. 26, 2020).

225 See U.S. Dep’t of Justice (Exec. Office for Immigration Review), Agency Decisions, https://www.justice.gov/eoir/ag-bia-decisions (last visited Nov. 26, 2020).

226 See, e.g., U.S. Dep’t of Justice (Exec. Office for Immigration Review), Board of Immigration Appeals, https://www.justice.gov/eoir/board-of-immigration-appeals (last visited Nov. 26, 2020); U.S. Patent & Trademark Office, Decisions, https://www.uspto.gov/patents-application-process/patent-trial-and-appeal-board/decisions (last visited Nov. 26, 2020); U.S. Citizenship & Immigration Servs., AAO Decisions, https://www.uscis.gov/administrative-appeals/aao-decisions (last visited Nov. 26, 2020).

227 The websites cited in the preceding footnote include useful explanations of the difference. See also, e.g., Nat’l Lab. Relations Bd., Cases and Decisions (Unpublished Board Decisions), https://www.nlrb.gov/cases-decisions/decisions/unpublished-board-decisions (last visited Nov. 30, 2020).

228 See Envtl. Prot. Agency (Env’t Appeals Bd.), supra note 223; see also, e.g., Equal Emp’t Opportunity Comn’n, Digest of Equal Employment Opportunity Law, https://www.eeoc.gov/ digest (last visited Nov. 30, 2020) (quarterly digest of recent cases); Equal Emp’t Opportunity Comm’n, Selected Noteworthy Federal Sector Appellate Decisions, https://www.eeoc.gov/ federal-sector/selected-noteworthy-federal-sector-appellate-decisions (last visited Nov. 30, 2020) (identifying and summarizing recent decisions of note).

229 See, e.g., U.S. Citizenship & Immigration Servs., AAO Practice Manual—About the Administrative Appeals Office Practice Manual, https://www.uscis.gov/administrative-appeals/aao-practice-manual (last visited Nov. 26, 2020).

48

e. At least three agencies (MSPB, NLRB, and BIA) in our study invite the public to subscribe to a listserv (through platforms like GovDelivery) to receive notices of new decisions.230 This strikes us an especially good practice that other agencies might consider adopting.

f. Not all agencies include materials from which the public can ascertain the appellate program’s legal authority. This is most commonly a problem when the authority derives from a delegation/assignment from an agency head that resides in some sort of unpublished rule (sometimes styled an “order”). Identifying and linking to such a document is a good practice. Better still is the practice, followed by some agencies, of providing for or referencing the delegation of authority in its CFR-codified rules of practice.231

2. Notice and Identification of Judicial Opinions. A pending ACUS project addresses litigation materials, including decisions on agency websites. One best practice, not specifically addressed there, we suggest for consideration is that, when posting an appellate decision on their websites, agencies should note any court decisions reviewing it. The NLRB accomplishes this indirectly, in an exemplary way, by including in its online docket each case’s subsequent judicial materials (e.g., petitions for review, motions, orders, and opinions).232

3. Appellate Proceedings: Oral Argument. It is especially important that appellate oral argument, which may address legal and policy issues of concern beyond the litigants to particular cases, be presumptively open to the public. (Of course, the Sunshine Act requires that boards and commissions open argument to the public.233) A few agencies’ rules require that proceedings be open.234 When argument is conducted by video, agencies should, whenever practicable, allow public access by phone or, better still, audio or video access.235 (Again, board and

230 U.S. Merit Sys. Prot. Bd., Of Note…—Subscribe to the MSPB Decision ListServ,

https://www.mspb.gov/decisions/casereports.htm (last visited Nov. 26, 2020) (left side of page); U.S. Dep’t of Justice (Exec. Office for Immigration Review), Agency Decisions, https://www.justice.gov/eoir/ag-bia-decisions (last visited Nov. 26, 2020); Nat’l Lab. Relations Bd., Weekly Summary, https://www.nlrb.gov/cases-decisions/decisions/weekly-summaries-decisions (last visited Nov. 30, 2020).

231 See, e.g., Equal Empl. Opportunity Comm’n, 29 C.F.R. § 1614.404(a). On the use of rules of practice, see Part III.A.2 supra.

232 See Nat’l Labor Relations Bd., Cases and Decisions, https://www.nlrb.gov/cases-decisions/cases (last visited Nov. 26, 2020).

233 See 5 U.S.C. § 552b(b). But see id. § 552b(c)(10) (exempting meetings concerning “disposition by the agency of a particular case of formal agency adjudication pursuant to the procedures in section 554 of this title or otherwise involving a determination on the record after opportunity for a hearing”).

234 See, e.g., Sec. & Exch. Comm’n, 17 C.F.R. § 201.201. 235 See, e.g., Fed. Mine Safety & Health Rev. Comm’n, Sunshine Act Meeting, 85 Fed. Reg.

60,786 (Oct. 8, 2020) (providing a call-in number).

49

commissions must do so under the Sunshine Act.) Live-streaming on the agency’s website would be ideal. Most agencies do not yet do so.

Some agencies provide advance notice of oral arguments prominently on their websites, usually in furtherance of their obligations under the Sunshine Act, sometimes voluntarily.236 Two agency websites we reviewed not only post such notices but also include especially accessible and well-written overviews (beyond Sunshine Act requirements) of the cases in which argument is to be held.237 Other agencies that hold argument on matters of concern beyond the litigants to a case might consider adopting this practice.

We did not encounter the practice of agencies posting on their websites transcripts, recordings, or videos of appellate oral arguments. Federal appellate courts increasingly post audio or video recordings of oral argument on their websites.238 It would be impracticable for most agency appellate programs that hold many oral arguments to do so in every case; and, volume aside, it would be of little value to the public for programs to do so in many routine cases. But some agencies might consider doing so in select cases that raise legal and policy issues of importance beyond the parties to the case. Posting at least some oral arguments online also may help parties, their lawyers and representatives, and the public more generally to understand how the oral argument process works at the agency and, in some cases, to better prepare for their own arguments before the agency.

4. Explanatory Materials for the Public. Most agency appellate programs post on their website at least some public-facing materials explaining what decisions they review and the procedures they use. Sometimes the explanations appear in free-standing documents that the user opens as a PDF, at other times as website text. Explanatory materials take any number of forms—e.g., practice guides, “questions and answers,” and fact sheets—and bear any number of titles. Some agencies post materials largely designed for internal use—such as practice manuals for enforcement staff239—that may also serve an explanatory function. (Most or all of these materials constitute “rules”240 under

.236 See, e.g., Envtl. Prot. Agency, Envtl. Appeals Bd., Upcoming Oral Arguments,

https://yosemite.epa.gov/oa/EAB_Web_Docket.nsf/Upcoming+Oral+Arguments?OpenView (last visited Nov. 30, 2020).

237 See Sec. & Exch. Comm’n, Sunshine Act Meeting, https://www.sec.gov/news/ openmeetings/2016/ssamtg032116.html (last visited Nov. 26, 2020); Occupational Safety & Health Review Comm’n, Rules—Guides, https://www.oshrc.gov/rules-guides (last visited Nov. 26, 2020).

238 See, e.g., U.S. Court of Appeals for the D.C. Circuit, Oral Arguments, https://www.cadc.uscourts.gov/internet/home.nsf/content/Oral+Arguments (last visited Nov. 26, 2020).

239 See, e.g., NAT’L LAB. RELATIONS BD., NLRB CASEHANDLING MANUAL, https://www.nlrb.gov/reports-guidance/manuals; see also Stack & Metzger, supra note 103, at 1241 & n.6 (describing the NLRB’s case-handling manual).

240 See 5 U.S.C. § 551(4).

50

the APA—though rules without the force of law in the sense that they are not judicially enforceable—but the terminology here is unimportant.)

Our overall impression is that these materials are generally well drafted and accessible to the intended audience. We are reluctant to pass any generalized judgment here because the intended audience for explanatory materials differs from agency to agency. For some agencies, the intended audience consists of mostly unrepresented parties; for others it consists almost exclusively of lawyers drawn from a specialized bar.

A few generalizations, though, are possible. First, the agency must take care to ensure that explanatory materials do not take the place of CFR-codified rules of practice and that the explanatory materials are not at variance with them. This is a more serious concern with hearing-level materials, but we have identified some suboptimal practices with appellate materials as well. They include the use of manuals that set forth certain rules of practice—to which both agency and litigant are seemingly bound—while disclaiming that they have any legal effect Second, explanatory materials are often most effective when they reference (and, better still, also link to) the underlying CFR-codified regulations, statutes, and other authoritative materials. That encourages litigants to cite the underlying source rather than the explanatory materials.

Third, some agencies have too many explanatory documents, with the result that information is presented in a fragmented, and sometimes redundant or even outdated, manner. Agencies should review these materials with an eye toward possibly simplifying, shortening, and consolidating them. And fourth, agencies should include as text on their website brief overviews of their appellate programs, even if they choose, as many do, to post comprehensive and lengthy explanatory documents. Here again the target audience is an important consideration.

5. Internal Decision-Making Processes. By internal decision-making processes, we contemplate such critical matters as how cases are assigned to adjudicators; what staff the program uses; what role staff play in screening and sorting cases, assigning them to different tracks (when agencies use tracking schemes), making recommendations about the disposition of appeals, and writing decisions; and the order in which cases are decided. Some agencies do not make this and other information about their internal decision-making processes available to the public.

One agency that does this especially well is the SSA Appeals Council. Its HALLEX manual effectively illustrates how information about internal decision-making processes can be presented.241 Another approach is to provide some explanation of internal processes in an agency’s rules of practice. Some rules of practice, for instance, explain what role agency staff plays. How much

241 See HALLEX, supra note 108, § I-2-0 through § I-2-10.

51

information is appropriate for a rule of practice depends on the particular program.

6. Information About Adjudicators. Locating information about appellate adjudicators other than agency heads—their method of appointment, terms and conditions of employment, permissible interactions with other agency officials, and so forth—is often more difficult than it should be. Such information about hearing-level adjudicators often resides only in unpublished agency policy documents. ACUS is addressing the subject of disclosure of information about adjudicators in a pending project.242 We refer readers to its associated report and committee-proposed recommendation.243

7. Case-Management Statistics. As best we can tell, all agencies maintain some case-management (or processing) statistics. Most, but not all, agencies share at least some data with the public, often through annual reports. Some agencies are clearly more forthcoming with data than others. What particular data are and should be disclosed to the public, and in what form, of course depend on the particular program and the program’s stated objective(s). We are reluctant to make cross-agency generalizations or offer best practices. But especially important data include case outcomes and case-disposition times as well as the number of dispositions and cases awaiting decision.

G. Common Challenges in Agency Appellate Programs During our interviews with the leaders of these agency appellate programs,

a half dozen challenges emerged as recurring themes. Not all agency officials expressed all of these concerns, and some of these issues have already been mentioned in Part III. But these concerns were common enough to merit some additional discussion here.

1. Volume and Backlogs. Just as it is a recurring theme in high-volume agency adjudication at the hearing level, many agency officials expressed concerns about the volume of cases they review and the backlogs of appeals that have accumulated. Some agency appellate programs are facing backlogs in the tens of thousands of appeals. Nearly all agency officials interviewed underscored the importance of managing the volume of appeals and striving to review the cases in an expeditious manner. For some agencies, however, the backlog just keeps growing due to the lack of sufficient personnel and resources to handle the heavy appellate docket. Some agency officials indicated that the backlogs may also be due in part to a more time-consuming de novo standard of review as well as, in some circumstances, an open-file policy that readily allows the appellant to supplement the administrative record under review.

242 Public Availability of Information About Agency Adjudicators, supra note 19. 243 See Schriever, supra note 130. The project also draws on Barnett, Non-ALJ Adjudicators,

supra note 16, published as Barnett & Wheeler, supra note 16; see also Barnett, supra note 130, at 563–82.

52

2. Resource Constraints. Many agency officials underscored that their appellate program lacks sufficient resources to effectively carry out its objectives. Not surprisingly, some agency appellate programs have more resources than others, sometimes for reasons other than case volume, and there seems to be a strong correlation between resources available at the hearing level and those available at the appellate level. When discussing resource constraints, many agency officials focused on human capital—in terms of both appellate adjudicators and staff attorneys and administrative personnel. But some officials also noted the lack of an electronic filing and case management system as well as insufficient technology for docket management, videoconference hearings, quality assurance, and developments in automation, machine learning, and artificial intelligence.

3. Vacancies, Turnover, and Personnel Law. In conjunction with their discussion of backlogs and resource constraints, many agency officials expressed frustration about the turnover in appellate adjudicators and staff. Some of these concerns are quite public, such as the lack of a decision-making quorum at the MSPB. Other incidences of vacancies and turnover escape public attention. Some agency officials suggested that personnel law plays a role in the delay in filling vacancies.

4. Electronic Case Management. A number of agency officials interviewed emphasized that their appellate programs still do not have an electronic case management system (eCMS). Instead, they deal with paper submissions and files, which leads to increased expense and decreased efficiency. Some of these appellate programs are in the process of transitioning to an eCMS. The agency officials with an eCMS emphasized the tremendous benefits in terms of not just cost and efficiency but also the ability to implement better quality assurance measures and experiment more fully with automation, machine learning, and artificial intelligence innovations.

5. Reconciling Competing Objectives. At the outset of each interview, we spent considerable time exploring the various reasons why the agency—by statute or regulation—has an appellate review program. Many agency officials interviewed, however, underscored that their appellate program faces competing objectives. Sometimes there are tensions in the objectives set forth in Part I.C between, for instance, error correction and a broader focus on inter-decisional consistency, or between policymaking and error correction. But far more often the competing objectives concern the tradeoffs between quality (i.e., engaging in thorough yet time-consuming review) and quantity (i.e., reducing backlogs and only correcting obvious errors). These tradeoffs were often mentioned in the interviews in the context of the parties’ ability to supplement the record and the standard of review that applies at the appellate stage.

6. Responding to Circuit-Court Decisions. Although not the central focus of our study and interviews, many agency officials noted in the interviews that their appellate adjudicators—and the agency as a whole—spent considerable time responding to judicial review. As further detailed in Part

53

III.C.11, the appellate adjudicators follow closely how their cases are being reviewed in the federal courts, especially in the circuit courts. Judicial remands to the agency force the agency to engage with the circuit-court’s reasoning and decide whether to embrace that reasoning as the agency’s own nationwide or only acquiesce in that particular circuit.

Aside from that circuit-court acquiescence question, the agency has the choice of how broadly or narrowly to construe the judicial command—or in the Chevron deference context with an ambiguous statute, whether to advance a different agency statutory interpretation than that embraced by the circuit court. There are also complicated questions about whether the agency head or appellate body should issue a precedential decision (or other informal guidance) in response to the judicial remand in order to help the hearing-level adjudicators incorporate the circuit-court command, or whether to just remand the case back to the hearing-level adjudicator to deal with all of those issues in the first instance on remand. The agency appellate programs in our study take different approaches to these judicial remand issues, and recognize that remands present both opportunities and challenges for their agency adjudication program more generally.

IV. RECOMMENDATIONS

Our recommendations draw from our findings (Part III), though not all findings resulted in a recommendation. We have been judicious in drawing recommendations from our findings given the preliminary nature of some of our findings and the heterogeneity of agency appellate programs. In considering these recommendations, agencies should of course weigh the costs and benefits of adoption in light of existing resources.

We organize the recommendations into six categories, which roughly correspond to the organization in Part III. The recommendations under each category are followed by a brief comment. Most of the comments track the findings in Part III. A. Objectives of Appellate Review

1. Agencies should identify what objective or objectives are served by their appellate programs, and they should design their processes and draft their procedural regulations accordingly.

2. Agencies should publicly announce—preferably by codifying them in procedural regulations—the objective or objectives of their appellate program.

As emphasized in this Report, appellate review can serve a number of different, and sometimes competing, objectives, ranging from error correction to policy formulation. An important question every agency should answer with respect to each appellate program is: What purpose or purposes does it serve? The answer to that question will necessarily underlie many of the procedural

54

choices an agency must make. They include, most obviously, the criteria for deciding which hearing-level decisions to review when review is discretionary. Having identified the purpose of a particular appellate program, the agency should identify it in its procedural regulations. We say “identify” rather than, say, “decide” because the objective of an agency appellate program may be dictated by statute.

B. Procedures for Appellate Review 1. Agencies should address all significant procedural matters governing

agency appellate review in procedural regulations—often styled as a “rules of practice” or “rules of procedure”—published in the Code of Federal Regulations rather than relegating them to non-legislative rules or other documents. Significant procedural matters pertaining to agency appellate review include:

(a) the availability of interlocutory review; (b) the procedures for initiating review; (c) the standards for granting review, if review is discretionary; (d) the scope and standard of review; (e) the allowable and required submissions by litigants—including

petitions, motions, and briefs—and their required contents; (f) the procedures and criteria for designating decisions as

precedential and the legal effect of such designations; (g) the record on review and the opportunity, if any, to submit new

evidence; (h) the availability of and criteria for oral argument and amicus

participation; (i) the availability of, criteria for, and procedures for reconsideration; (j) in the case of multi-member appellate boards, councils, and the

like, the authority to assign decision-making authority to fewer than all members (e.g., panels); and

(k) any administrative exhaustion requirements that must be satisfied before seeking judicial review.

2. Agencies should consider including in the procedural regulations governing their appellate programs: (a) a brief statement or explanation of each program’s review authority, structure, and decision-making components; and (b) for each provision based on a statutory source, an accompanying citation to that source.

3. When revising existing or adopting new appellate rules, agencies should review the appellate rules (Rules 400-450) in the Administrative Conference’s Model Rules of Agency Adjudication (rev.

55

2018) (Appendix B to this Report) in deciding what the rules should provide.

4. When adopting new or materially amending existing procedural regulations, agencies should voluntarily use notice-and-comment procedures or other mechanisms for soliciting public input, notwithstanding the procedural rules exemption of 5 U.S.C. § 553(b)(A), unless the costs clearly outweigh the benefits of doing so.

ACUS has recommended that hearing-level adjudication programs include a “complete statement of important procedures”—that is “all important procedures and practices that affect persons outside the agency in procedural regulations”—in “procedural regulations” published in the Federal Register and Code of Federal Regulations.244 This should undoubtedly apply to appellate programs as well. (We refer to “procedural regulations” as “CFR-codified rules of practice.”) Codifying important procedures serves several objectives, not least the rule-of-law objective of laying down in explicit terms the procedures to which both the agency and litigants must follow in both agency proceedings and the federal courts.

No hard-and-fast rule dictates which procedural matters should be addressed in CFR-codified procedural regulations and which may be addressed instead in guidance documents like practice manuals. We include above a non-exhaustive list drawn in part from the appellate provisions of ACUS’s Model Adjudication Rules (rev. 2018) (Appendix B), to which we commend all agencies when reviewing their rules of practice.

Consistent with ACUS’s long-standing position,245 we recommend that agencies voluntarily solicit public comment on any proposed procedural rules subject to exemption of 5 U.S.C. § 553(b)(A), unless the costs of doing so clearly outweigh the benefits. Rarely will the costs exceed the benefits. Of course, some changes to procedure rules will be so minor and technical as to not call for public comment. We assume that point need not be made explicit in a recommendation.

C. Case Selection for Appellate Review 1. Based on the agency-specific objectives for appellate review, agencies

should consider whether review should be mandatory or discretionary (assuming they have statutory authority to do so); if discretionary, the standards for granting review should track the purposes of the appellate program, and they should be published in the procedural regulations.

2. Agencies should consider implementing procedures for sua sponte appellate review of non-appealed hearing-level adjudications as well

244 Recommendation 2016-4, supra note 30, § 28 at 94,316. 245 See, e.g., Recommendation 92-1, supra note 121, § 2.

56

as for hearing-level adjudicators to refer cases to the appellate body for review.

As documented in Part III.C of this Report, some agency appellate programs have mandatory review, and others use a discretionary review system. This design decision should be informed by the agency’s specific objectives for appellate review. As discussed in Part I.D, prior ACUS recommendations encouraged agencies to consider embracing discretionary review, especially if they face high volumes of appeals and are more concerned with inter-decisional consistency or policymaking instead of mere error correction. For agencies that currently have or decide to adopt a discretionary review process, they should publish criteria for such discretionary review—criteria that should reflect agency-specific objectives for appellate review.

Especially if agencies are concerned about inter-decisional consistency, they should consider embracing mechanisms for sua sponte appellate review of non-appealed hearing-level adjudications. As detailed in Part III.C, the SSA and SEC have both implemented such review as a form of quality assurance. Agency appellate programs concerned with policymaking or even just inter-decisional consistency may also want to consider implementing procedures for hearing-level adjudicators to refer cases to the appellate program for review.

D. Appellate Decision-Making Processes and Decisions 1. Whenever possible, agencies should maintain a single or integrated

electronic case management system (eCMS) for both hearing-level and appellate programs, or otherwise design their eCMS to ensure that hearing records are easily accessible to appellate adjudicators.

2. Agencies should explore ways to implement screening or sorting methods for appeals and, where appropriate, the grouping of appeals on the merits based on difficulty, common legal issue, subject matter, record length, or other relevant factors in order to better leverage adjudicator expertise and economies of scale.

3. Agencies should consider how to better use staff attorneys and law clerks at both the screening and merits stages.

4. Agencies should consider using automation, artificial intelligence, and machine learning to assist in screening and sorting appeals.

5. Agencies should set their scope and standard of review to be consistent with agency-specific objectives for their appellate program. For most appellate programs, it is not advisable to have a de novo standard of review. Nor is it prudent to have a de novo scope of review where appellants can freely introduce any new evidence on appeal.

6. Agencies should strive to improve the readability and overall quality of their appellate decisions, including an emphasis on plain language

57

and experimentation with decision templates and other quality-improving measures.

7. Agencies should establish clear criteria and processes for selecting and publishing precedential opinions, especially for appellate programs with objectives of policymaking or inter-decisional consistency.

8. Agencies should assess the value of oral argument and amicus participation in their appellate program and should establish clearer rules and criteria for both.

As detailed in Part III.D, some agency appellate programs face serious challenges in overloaded dockets and delays in the appellate decision-making process. The recommendations detailed above are all based on best practices uncovered in the agency interviews and appellate program overviews. Among other things, many agency officials observed that it is important that the appellate body have the appropriate scope and standard of review, in order to focus limited resources on the agency’s appellate review objectives. Many offered that a trial de novo scope of review was inconsistent with the purpose of their agency appellate structures, and the same with a de novo standard of review.

The final three recommendations above aim at improving the quality of the appellate decision-making process and the quality of the appellate decisions themselves. As discussed in Part III.C, many agency appellate programs do not have effective rules, criteria, and processes with respect to oral argument, amicus briefs, and precedential opinions. Other agencies, by contrast, have much more developed rules and processes that could serve as a model.

Similarly, our review of appellate decisions uncovered a great diversity in terms of opinion structure, format, style, readability, and overall quality. Many agencies have dedicated considerable resources to improve the quality of appellate decisions, including experimentation with decision templates and standard language and the use of staff attorneys, paralegals, and law clerks. Agencies should strive to make their appellate decisions more accessible to the appellants in the program—many of whom have no legal representation.

E. Administration, Management, and Bureaucratic Oversight 1. Agency appellate programs should promptly transmit their

precedential decisions to all program adjudicators and, directly or through hearing-level programs, to hearing-level adjudicators (as appropriate). Appellate programs should include in their transmittals, when feasible, summaries and explanatory materials.

2. Agency appellate programs should notify their adjudicators of significant federal-court decisions reviewing the agencies’ decisions and, when providing notice, explain the significance of those decisions to the program. As appropriate, agencies should notify adjudicators of any policies governing whether and when they will acquiesce in the decisions of the federal courts of appeals.

58

3. Agency appellate programs where decision making relies extensively on their own precedential decisions should consider preparing indexes and digests—with annotations and comments, as appropriate—to help adjudicators identify those decisions and their significance.

4. Agency appellate programs should regularly communicate with agency rulewriters and other agency policymakers—and, as appropriate, institutionalize communication mechanisms—to address whether interpretations and policies addressed in their decisions should be addressed by rule rather than case-by-case adjudication. Appellate programs should also communicate to agency policymakers, congressional liaisons, and other appropriate officials any needed statutory amendments that the program may identify.

5. The Office of the Chairman of the Administrative Conference should provide for, as authorized by statute, the “interchange among administrative agencies of information potentially useful in improving” (5 U.S.C. § 594(2)) agency appellate programs. The subjects of interchange might include electronic case management systems, procedural innovations, quality-assurance reviews, and common management problems.

Our findings from Part III.E have led us to several best practices on this subject that are applicable to most or if not all appellate programs. Many administrative reforms, of course, must be undertaken on a program-by-program basis to account for, among other things, differences in mission, size, and resources. Hence our selectivity.

In identifying the best practices above, we have been guided by the objectives of appellate review identified in Part I.C. Most or all of the recommendations serve one or more of objectives that nearly all programs share to varying degrees. We have not addressed matters addressed in other ACUS recommendations, such as the use of video technology, that have no special application to appellate programs.

F. Public Disclosure and Transparency 1. Agencies should disclose on their websites any rules (sometimes styled

as “orders”) by which an agency head has delegated review authority to appellate adjudicators.

2. Agencies should consider announcing, livestreaming, and maintaining audio or video recordings on their websites of appellate proceedings (including oral argument) that present significant legal and policy issues likely to be of interest to regulated parties and other members of the public. Brief explanations of the issues to be addressed by oral argument might usefully be included in website notices of oral argument.

59

3. Agencies should include on their websites brief and accessibly written explanations as to how their internal decision-making processes work and, as appropriate, include links to explanatory documents appropriate for public disclosure. Specific subjects agencies should consider addressing include: the process of assigning cases to adjudicators (when fewer than all adjudicators participate in a case), the role of staff, and the order in which cases are decided.

4. When posting decisions on their website, agencies should clearly distinguish between precedential and nonprecedential decisions. Agencies should also include a brief explanation of the difference.

5. When posting decisions on their website, agencies should consider, as practicable, including brief summaries of precedential decisions and, for precedential decisions at least, citations to court decisions reviewing them.

6. Agencies should include on their website any digests and indexes of decisions they maintain. It may be appropriate to remove any material exempt from disclosure under the Freedom of Information Act or other laws.

7. Agencies should affirmatively solicit suggestions for rule changes on its website.

Several ACUS recommendations already address public disclosure and transparency in adjudication programs. They include (a) Recommendation 2016-4, Evidentiary Hearings Not Required by the Administrative Procedure Act,246 which provides that adjudication hearings be presumptively open to the public (§ 18) and that agencies include on their website manuals, guides, and other explanatory materials for the benefit of agency officials and parties (§ 29); (b) Recommendation 2018-5, Agency Disclosure of Adjudication Rules,247 and Recommendation 2017-1, Adjudication Materials on Agency Websites,248 which together provide guidance to agencies on what rules, decisions, and other adjudication materials should be posted on agency websites and in what manner; and (c) a forthcoming recommendation, Public Availability of Information About Agency Adjudicators,249 that provides for the disclosure of specific categories of information from which the public can assess the relative impartiality and constitutional status of agency adjudicators. Another

246 Recommendation 2016-4, supra note 30. 247 Recommendation 2018-5, supra note 104. 248 Recommendation 2017-1, supra note 220. 249 Proposed Recommendation, Public Availability of Information About Agency

Adjudicators, supra note 130.

60

forthcoming ACUS recommendation, Agency Litigation Webpages,250 addresses the disclosure of court decisions arising from review of adjudication decisions.

These recommendations, for the most part, apply equally to hearing-level and appellate programs. We leave it to ACUS to decide whether they should be incorporated into the recommendation associated with this Report. All of these recommendations are focused on agency appellate programs and are consistent with the above-mentioned recommendations.

CONCLUSION

It has been nearly four decades since ACUS last examined agency appellate systems. This Report has covered extensive ground. In so doing, it has uncovered a number of common challenges. It has documented various best practices that agencies have developed to improve their appellate programs. Ultimately, however, this Report is exploratory in nature. It perhaps raises more questions than provides answers.

This Reports surely will not be the last word on the subject. As flagged throughout the Report, our study has uncovered a number of issues that merit further exploration by ACUS, agencies, policymakers, and scholars. For example, the subject of the readability and overall quality of appellate decisions merits further inquiry. So does the function and process of creating precedential appellate decisions. Indeed, much more attention should be paid to assessing the effectiveness of appellate review as a quality-assurance mechanism, compared to other potential approaches to quality assurance. The use of automation, artificial intelligence, and machine learning in appellate review will no doubt spark much more scholarly and policy attention.

Looking outside of the agency appellate review system itself, this Report highlights numerous issues that merit further attention. For instance, the interaction between hearing-level and appellate adjudicators has barely been studied, yet that relationship is critical for an agency to realize its objectives for administrative adjudication. The relationship between agency adjudicators and federal courts is similarly understudied. And even less attention has been paid to the interaction between appellate adjudicative bodies and agency rulewriters, congressional liaisons, and the agency as a whole.

In sum, when it comes to agency appellate review, much more work needs to be done—empirically, doctrinally, and theoretically. With how important agency appellate review has become for administrative adjudication, we highly doubt that it will be another four decades before ACUS returns to the subject.

250 Admin. Conf. of the U.S., Agency Litigation Webpages, https://www.acus.gov/research-

projects/agency-litigation-webpages (last visited Nov. 26, 2020).

APPENDIX A TEMPLATE FOR AGENCY-SPECIFIC OVERVIEWS

Appendices C–N, more than two hundred pages in length, reproduce the overviews of the twelve study agencies. These appendices appear on the ACUS website at https://www.acus.gov/research-projects/agency-appellate-systems.

The template we used for the overviews follows on the next page.

ii

Administrative Conference of the United States

Agency Appellate programs

Christopher W. Walker Matthew Lee Wiener

[AGENCY NAME]

INTRODUCTION

[narrative overview of program]

CHARACTERISTICS OF SYSTEM

(as ascertainable from public sources) Governing Law Procedural Law

Substantive Law

Miscellaneous

Hearing-Level Proceedings in Appellate program Hearing-Level Decisions Appealable

Nature of Hearing-Level Proceedings

Nature of Hearing-Level Decision

Transfer of Case to Appellate Body After Hearing-Level Decision

iii

Miscellaneous

Identity of Reviewing/Appellate Authority and Its Legal Status Reviewing Authority(ies)

Legal Status of Reviewing Authority(ies)

Miscellaneous

Institutional Attributes of Appellate/Reviewing Authority(ies) Number of Members

Qualification Requirements

Party Affiliation Requirement in Appointment

Method of Appointment

Term of Appointment

Statutory Removal Protections

Location within Agency; Basis of Legal Authority

Authority to Delegate to Subunit(s); Designating Official and Process

Quorum Requirement

Authority and Function of Appellate Authority’s Head

iv

Internal Management Structure of Appellate Authority

Miscellaneous

Nature, Form, and Timing of Appeal Nature of Appeal: Discretionary Versus as of Right; If Discretionary, Standards for Allowance

How Appeal Initiated

Time For Appealing

If No Appeal Taken from Hearing Officer’s Decision

If Appeal Taken

Review of Decisions on Own Initiative (Without Request of a Party)

Miscellaneous

Appellate Authority’s Procedures Record on Review

Submissions by Parties in Support of Appeal

Issue Preservation

Open or Closed Record on Appeal? Submission and Consideration of New Evidence on Appeal

v

Standard of Review

Consultation with Staff and Other Agency Officials

Oral Argument

Amicus Participation; Intervention; etc.

Public Access to Hearings

Staff’s Role in Writing Decisions

Deadlines for Decision

Nature of Decision

Reconsideration, Rehearing, etc.

Miscellaneous

Other Case-Management Features Interlocutory Appeals: Availability, Procedures, Standard

Assignment of Cases

Special Case-Selection Techniques (e.g., Artificial Intelligence) When Appeal Is Not as of Right

N/A

Aggregation

Miscellaneous

vi

Form of Decisions, Publication, and Precedential Status

Form of Decision

Signed or Per Curiam

Dissents

Publication

Where Published

Precedential Status

Miscellaneous

Extra-Adjudicative Activities of Appellate Authority to Direct or Review Activities of Adjudicators Below Guidance Documents Governing Hearing-Level Adjudicators

Feedback to Adjudicators

Quality-Assurance Reviews and Related Mechanisms

Participation of Appellate Body in Substantive Rulemaking

Miscellaneous

Miscellaneous Alternative Dispute Resolution (ADR)

vii

Participation of Appellate Body in Agency Decisions on Judicial Review

Role and Participation of Appellate Body in Writing Rules

Miscellaneous

viii

APPENDIX B APPELLATE PROVISIONS OF MODEL ADJUDICATION RULES

(RULES 400–450)

ADMINISTRATIVE REVIEW

Rule 400. Interlocutory Review (A) Interlocutory review should be handled on an expedited basis. (B) A party that seeks interlocutory review of an adjudicator’s decision, or

part thereof, must file a petition with the Adjudicator. The petition must: (1) be filed with the Adjudicator within __ days after the Adjudicator’s

decision; (2) designate the decision (or part thereof) from which review is sought;

and (3) set forth the grounds on which review is sought, including all

applicable points of fact and law, and the reasons why interlocutory review is warranted under Rule 400(D).

(C) Any party that opposes the petition may file a response within __ days after service of the petition.

(D) The Adjudicator must certify the ruling for interlocutory review by [the Agency] if the Adjudicator determines that: (1) the decision involves a controlling question of law about which there

is substantial ground for difference of opinion; and (2) an immediate review will materially advance the completion of the

adjudication, or subsequent review by [the Agency] will provide an inadequate remedy.

(E) Within __ days after the Adjudicator’s ruling on a petition to certify a decision under Rule 400(B), the petitioner may apply to [the Agency], whether or not the Adjudicator has certified the decision under Rule 400(D), to allow the interlocutory review sought in the petition. The application must reference the petition filed under Rule 400(B), all filings made with the Adjudicator in support of or in opposition to the petition, and the Adjudicator’s decision on the petition. The application must not otherwise set forth the grounds on which interlocutory review is sought or contain any argument, unless [the Agency] orders otherwise. No response to any application made under this subsection may be permitted unless [the Agency] orders otherwise.

(F) Any petition or application filed under this rule may be summarily dismissed whenever the Adjudicator or [the Agency], respectively, determines that review is not appropriate.

ix

(G) [The Agency] may, on its own motion, certify an order for interlocutory review under this rule in its discretion.

(H) If [the Agency] decides to allow interlocutory review, [the Agency] must decide the matter on the basis of the administrative record and briefs submitted to the Adjudicator, without further briefs or oral argument, unless [the Agency] orders otherwise.

(I) The filing of an application for interlocutory review and the certification of a ruling for interlocutory review does not stay proceedings before the Adjudicator unless (s)he or [the Agency] so orders.

Official Comment

1. An agency may wish to consider whether to provide interlocutory review in particular situations (e.g., an adverse ruling on a motion for the Adjudicator’s recusal, a ruling suspending an attorney from participation in the adjudication, a ruling denying or terminating intervention or limited participation, or a ruling requiring the production of information claimed to be privileged) even if the Adjudicator has denied the petition. In such instances, the agency may also wish to consider whether to provide by rule that if it does not reverse the Adjudicator’s denial of the application for interlocutory review within __ days, the Adjudicator’s ruling is affirmed. See Admin. Conf. of the U.S., Recommendation 71-1, Interlocutory Appeal Procedures, 38 Fed. Reg. 19,787 (July 23, 1973).

2. The approach taken in this rule is largely consistent with ACUS Recommendation 71-1. The rule’s two-step process—Adjudicator certifies decision, Agency then decides whether to take review—tracks the general statute governing interlocutory appeals of orders of the federal district courts. See 28 U.S.C. § 1292(b). Specific types of district-court orders may be subject to other rules. See, e.g., Fed. R. Civ. P. 23(f) (allowing a court of appeals to permit an appeal of a district-court order granting or denying a class-action certification whether or not the district court has certified the order for appeal). The rule departs from § 1292(b) in allowing an Agency to hear interlocutory review even if the Adjudicator does not certify the decision for review under Rule 400(D). It is anticipated that interlocutory review in the absence of certification will be permitted only in the narrowest circumstances.

Rule 410. Petitions for Review

(A) Any party may file with [the Agency] a petition for review within __ days after issuance of the Adjudicator’s decision. Two or more parties may join in the same petition.

(B) A petition for review, no more than __ words, must be filed only upon one or more of the following grounds:

x

(1) a finding of material fact is not supported by substantial evidence; (2) a necessary legal conclusion is erroneous; (3) the decision is contrary to law or to the duly promulgated rules or

decisions of [the Agency]; (4) a substantial question of law, policy, or discretion is involved; or (5) a prejudicial error of procedure was committed.

(C) Each issue must be plainly and concisely stated and must be supported by citations to the administrative record when assignments of error are based on the administrative record, and by statutes, regulations, cases, or other principal authorities relied upon. Except for good cause shown, no assignment of error by any party may rely on any question of fact or law not presented to the Adjudicator.

(D) A statement in opposition to the petition for review may be filed, within __ days after the date on which petitions are due.

(E) Review by [the Agency] is not a matter of right, but within the sound discretion of [the Agency]. A petition not granted within __ days after the issuance of the Adjudicator’s decision is deemed denied.

(F) [The Agency], at any time within __ days after the issuance of the Adjudicator’s decision, may review the decision on its own authority.

(G) A petition for review under this section is, under 5 U.S.C. § 704, a prerequisite to the seeking of judicial review of the final agency action. Unless [the Agency] provides otherwise, the effect of filing a petition for review is to stay the decision of the Adjudicator.

Official Comment

1. (to subsection (A)): Depending on agency statute or regulations, adjudicators ordinarily issue either initial decisions or recommended decisions. Initial decisions become effective as the agency’s decision unless a party seeks review or the agency, on its own initiative, elects to review the decision. Recommended decisions do not go into effect without further agency action and are issued in those cases where the agency will automatically review the decision. See 5 U.S.C. § 557(b). Agencies ordinarily have somewhat different procedures for review of initial and recommended decisions. For the purpose of filing petitions for agency review, this rule is limited to initial decisions. In the case of recommended decisions, however, an agency must take affirmative action to provide for review (such as by directing the filing of exceptions and briefs) and to render a final decision (such as by issuing its own decision or affirming the Adjudicator’s recommended decision).

xi

2. (to subsection (A)): ACUS has recommended that agencies establish an administrative review regime that limits the scope of agency review of decisions of Adjudicators in routine cases but authorizes agencies, on their own authority or upon request of a party, to review significant questions of policy, fact, procedure, or discretion fully as if the agency were making an initial decision. See Admin. Conf. of the U.S., Recommendation 68-6, Delegation of Final Decisional Authority Subject to Discretionary Review by the Agency, 38 Fed. Reg. 19,783 (July 23, 1973). Some statutes accord parties an automatic right of review by the agency. In such circumstances, agencies must modify this rule to provide for automatic review. See, e.g., 29 C.F.R. §§ 102.45, 102.46 (NLRB).

3. (to subsection (E)): In the interest of encouraging prompt appellate review of an adjudicator’s decision, this rule provides that petitions for discretionary review that are not granted within a period of time are deemed denied. Alternatively, agencies may elect in their regulations to stay an initial decision automatically any time a petition for discretionary review is filed until such time as the agency has disposed of the petition. “Effectiveness” of a decision is not necessarily the same as “finality.” See Comment 1, Rule 440 (Final Decision).

4. (to subsection (G)): The APA allows agencies to require an adversely affected party to ask a “superior agency authority” to review a subordinate agency decision before going to court, provided that subordinate decision is “inoperative,” i.e., not final, while the party is seeking review by the superior agency authority. 5 U.S.C. §§ 557(b), 704. In Darby v. Cisneros, 509 U.S. 137 (1993), the Supreme Court held that, when agency regulations simply authorize—but do not require—a party to seek administrative review, a party does not fail to exhaust required administrative remedies by foregoing the option of seeking administrative review. Rule 410(G) explicitly provides that, unless the agency provides otherwise, the filing of a petition for review is an administrative prerequisite to filing a petition for judicial review. This approach keeps ultimate decisional responsibility with the agency and avoids judicial review of issues on which the agency has not had an opportunity to rule. In some cases, an agency may wish to provide that exhaustion of administrative remedies is not a prerequisite to judicial review. In addition, there may be situations where an agency may choose to allow the Adjudicator’s decision to become operative pending administrative review by a superior administrative authority. In these latter cases, the Adjudicator’s decision is immediately judicially reviewable.

Rule 411. Record Before the Agency [The Agency] must decide each matter on the basis of the whole administrative record.

xii

Rule 412. Additional Evidence

(A) Upon its own motion or the motion of any party, [the Agency] may allow the submission of additional evidence.

(B) A party may file a motion for leave to adduce additional evidence before the issuance of a decision by [the Agency]. The motion must show with particularity that such additional evidence is material and that there were reasonable grounds for failure to adduce such evidence previously.

(C) [The Agency] may, as appropriate, accept or hear additional evidence, or remand or refer the proceedings to the Adjudicator for the taking of additional evidence.

Official Comment

1. 5 U.S.C. § 557(b) provides, “On appeal from or review of the initial decision, the agency has all of the powers which it would have in making the initial decision except as it may limit the issues on notice or by rule.”

Rule 420. Appellate Briefs

(A) Unless [the Agency] orders otherwise, a party must file a brief in support of its petition for review within __ days after [the Agency] grants the petition. If a petitioner fails to file a timely brief, the order granting review may be vacated. Other parties may file any briefs they wish considered by [the Agency] within __ days after the petitioner’s brief is served. If [the Agency] orders review on its own motion, all parties must file any briefs they wish considered by [the Agency] within __ days of the order, unless the Agency otherwise orders.

(B) Except by permission of [the Agency], a brief must not exceed __ words.

Official Comment 1. The APA accords an agency on review of an adjudicator’s decision “all the

powers which it would have in making the initial decision except as it may limit the issues on notice or by rule.” 5 U.S.C. § 557(b). Nonetheless, agencies typically do not review every issue decided by an adjudicator. Because agencies (like courts) ordinarily impose limitations on the length of appellate briefs, parties must be selective about the issues they raise for appellate review.

2. Agencies may adapt these regulations to provide that briefs supporting the petition for review must be filed on the same date and responsive briefs

xiii

filed later. Agencies may also authorize the filing of reply briefs in appropriate cases.

Rule 421. Amicus Briefs

(A) Any agency of the United States or limited participant in the proceedings below may file an amicus-curiae brief with [the Agency]. Any other amicus curiae may file a brief only with the consent of [the Agency] after the filing of a motion under Rule 421(B) or with the consent of the parties and a statement to that effect included with the brief.

(B) A motion for leave to file must accompany the brief and state (1) the movant’s interest in the adjudication, and (2) the reason why an amicus brief is desirable and why the matters

asserted are relevant to the disposition of the case. (C) Any amicus brief filed by anyone (other than an agency of the United

States), including by limited participants in the proceedings below, must state whether a party’s counsel authored the brief in whole or in part, a party or a party’s counsel contributed money that was intended to fund preparing or submitting the brief, and a person—other than the amicus curiae, its members, or its counsel—contributed money that was intended to fund preparing or submitting the brief and, if so, identify each such person.

(D) Except by permission of [the Agency], an amicus-curiae brief may not exceed __ words.

(E) An amicus curiae must file its brief, accompanied by a motion when necessary under these rules, no later than 7 days after the party which amicus curiae supports files its principal brief. An amicus curiae that does not support either party must file its brief no later than 7 days after the brief filed by the last filing party.

(F) Except by [the Agency’s] permission, an amicus curiae may not file a reply brief.

Official Comment

1. This rule largely draws from FRAP 29. Rule 430. Oral Argument [The Agency] may permit oral argument in its discretion. The order scheduling a case for oral argument must contain the allotment of time for each party and order of presentation for oral argument before [the Agency].

xiv

Official Comment 1. Oral argument is not a mandatory part of the appellate process. It is

designed to permit an agency to address issues that may have been left unresolved in the briefs or about which agency members have questions.

Rule 440. Final Decision

(A) If no petition for review is filed, and [the Agency] has not taken review of the Adjudicator’s decision on its own authority, the decision becomes the final decision of [the Agency] __ days after issuance.

(B) When a case stands submitted for final decision on the merits, [the Agency] will dispose of the issues presented by entering an appropriate order, which will include findings and conclusions and the reasons or bases therefor. In appropriate cases, [the Agency] may simply adopt the Adjudicator’s decision.

Official Comment

1. Effectiveness and finality of a decision are not necessarily synonymous concepts. Compare Comment 4 (to subsection (G)), with Rule 410 (Petitions for Review).

Rule 450. Reconsideration

(A) Any party may file a motion for reconsideration of a final order issued by [the Agency].

(B) Unless the time is shortened or enlarged by [the Agency], motions for reconsideration must be filed within __ days after service of the final order issued by [the Agency].

(C) A motion for reconsideration must be no more than __ words and must state, briefly and specifically, the matters of record alleged to have been erroneously decided, the ground(s) relied upon, and the relief sought. No responses to motions for reconsideration may be filed unless requested by [the Agency].

Official Comment

1. Agencies that do not want to provide for reconsideration should not adopt this rule.

2. Some statutes make the filing of a petition for rehearing or reconsideration with an agency a jurisdictional prerequisite for judicial review. See, e.g., 15 U.S.C. § 717r (Natural Gas Act). Many agencies allow petitions for reconsideration as a matter of discretion. This rule provides for motions

xv

for reconsideration of final agency decisions. Because reconsideration is intended to be an exceptional remedy, the rule provides that no responses to motions for reconsideration may be filed unless requested by the agency. This rule follows the approach taken in FRAP 35 and 40 with respect to petitions for rehearing and suggestions of rehearing en banc.

xvi

INDEX OF ONLINE APPENDICES C–N: AGENCY-SPECIFIC OVERVIEWS

Appendix C Department of Health and Human Services (HHS):

Departmental Appeals Board—Medicare Appeals Council

Appendix D

Department of Homeland Security (DHS), U.S. Citizen and Immigration Services: Administrative Appeals Office (USCIS)

Appendix E Department of Justice (DOJ): Board of Immigration Appeals (BIA)

Appendix F Department of Labor (DOL): Administrative Review Board

Appendix G Department of Veterans Affairs (VA): Board of Veterans Appeals (BVA)

Appendix H Environmental Protection Agency (EPA): Environmental Appeals Board (EAB)

Appendix I Equal Opportunity Employment Commission (EEOC)

Appendix J Merit Systems Protection Board (MSPB)

Appendix K National Labor Relations Board (NLRB)

Appendix L Patent and Trademark Office (PTO): Patent Trial and Appeals Board (PTAB)

Appendix M

Securities and Exchange Commission (SEC)

Appendix N Social Security Administration (SSA): Appeals Council